| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 371900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.0873817205429077, | |
| "learning_rate": 3.125e-06, | |
| "loss": 7.5939, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.8078175187110901, | |
| "learning_rate": 6.25e-06, | |
| "loss": 5.8309, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8323305249214172, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 5.3685, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.9240570068359375, | |
| "learning_rate": 1.25e-05, | |
| "loss": 5.158, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.988000750541687, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 5.0109, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.1597453355789185, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 4.8668, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.1680865287780762, | |
| "learning_rate": 2.1875e-05, | |
| "loss": 4.759, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.106913447380066, | |
| "learning_rate": 2.5e-05, | |
| "loss": 4.656, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.177737832069397, | |
| "learning_rate": 2.8125000000000003e-05, | |
| "loss": 4.5671, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.0818839073181152, | |
| "learning_rate": 3.125e-05, | |
| "loss": 4.4873, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.1852740049362183, | |
| "learning_rate": 3.4371875e-05, | |
| "loss": 4.4182, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.0693737268447876, | |
| "learning_rate": 3.7496875e-05, | |
| "loss": 4.3542, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.0483555793762207, | |
| "learning_rate": 4.061875e-05, | |
| "loss": 4.2985, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.0645984411239624, | |
| "learning_rate": 4.374375e-05, | |
| "loss": 4.2398, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.0418798923492432, | |
| "learning_rate": 4.686875e-05, | |
| "loss": 4.1925, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.9964326620101929, | |
| "learning_rate": 4.999375e-05, | |
| "loss": 4.1419, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.0009121894836426, | |
| "learning_rate": 5.3115625000000005e-05, | |
| "loss": 4.0911, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.9670995473861694, | |
| "learning_rate": 5.6240625e-05, | |
| "loss": 4.0554, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.3076399445722965, | |
| "eval_loss": 4.27287483215332, | |
| "eval_runtime": 153.7758, | |
| "eval_samples_per_second": 376.659, | |
| "eval_steps_per_second": 5.892, | |
| "step": 18595 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.033874750137329, | |
| "learning_rate": 5.93625e-05, | |
| "loss": 4.015, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.0117688179016113, | |
| "learning_rate": 6.24875e-05, | |
| "loss": 3.9733, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.9974212646484375, | |
| "learning_rate": 6.56125e-05, | |
| "loss": 3.9327, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.0678539276123047, | |
| "learning_rate": 6.8734375e-05, | |
| "loss": 3.8969, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.0113184452056885, | |
| "learning_rate": 7.185625e-05, | |
| "loss": 3.8656, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.955341637134552, | |
| "learning_rate": 7.498125e-05, | |
| "loss": 3.8339, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.9914633631706238, | |
| "learning_rate": 7.8103125e-05, | |
| "loss": 3.7996, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.9566612839698792, | |
| "learning_rate": 8.122500000000001e-05, | |
| "loss": 3.7732, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.9543666839599609, | |
| "learning_rate": 8.435e-05, | |
| "loss": 3.751, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.895650327205658, | |
| "learning_rate": 8.7471875e-05, | |
| "loss": 3.7278, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.923766553401947, | |
| "learning_rate": 9.0596875e-05, | |
| "loss": 3.7032, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.8873268365859985, | |
| "learning_rate": 9.3721875e-05, | |
| "loss": 3.683, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.8692554235458374, | |
| "learning_rate": 9.684062500000001e-05, | |
| "loss": 3.6674, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.8744521737098694, | |
| "learning_rate": 9.9965625e-05, | |
| "loss": 3.6463, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.8577896952629089, | |
| "learning_rate": 9.970903206825538e-05, | |
| "loss": 3.6334, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.8577241897583008, | |
| "learning_rate": 9.941512209473375e-05, | |
| "loss": 3.6116, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.8745716214179993, | |
| "learning_rate": 9.912091791703443e-05, | |
| "loss": 3.5939, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.8160855770111084, | |
| "learning_rate": 9.882671373933511e-05, | |
| "loss": 3.5793, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.8619437217712402, | |
| "learning_rate": 9.853250956163578e-05, | |
| "loss": 3.5589, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.36337055631321796, | |
| "eval_loss": 3.745145320892334, | |
| "eval_runtime": 155.2906, | |
| "eval_samples_per_second": 372.985, | |
| "eval_steps_per_second": 5.834, | |
| "step": 37190 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.8179039359092712, | |
| "learning_rate": 9.823830538393646e-05, | |
| "loss": 3.5228, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.7979289293289185, | |
| "learning_rate": 9.794439541041483e-05, | |
| "loss": 3.5131, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 0.8003087043762207, | |
| "learning_rate": 9.765019123271551e-05, | |
| "loss": 3.4951, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.805997908115387, | |
| "learning_rate": 9.735657546337158e-05, | |
| "loss": 3.4889, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.8201219439506531, | |
| "learning_rate": 9.706237128567226e-05, | |
| "loss": 3.4838, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 0.7858601808547974, | |
| "learning_rate": 9.676816710797295e-05, | |
| "loss": 3.4712, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 0.7911462187767029, | |
| "learning_rate": 9.647396293027361e-05, | |
| "loss": 3.4644, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.7916685342788696, | |
| "learning_rate": 9.61797587525743e-05, | |
| "loss": 3.454, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 0.7922804951667786, | |
| "learning_rate": 9.588584877905267e-05, | |
| "loss": 3.4491, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 0.7632368803024292, | |
| "learning_rate": 9.559164460135335e-05, | |
| "loss": 3.439, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.7886542677879333, | |
| "learning_rate": 9.529773462783173e-05, | |
| "loss": 3.4333, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.7786866426467896, | |
| "learning_rate": 9.50035304501324e-05, | |
| "loss": 3.426, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 0.760708749294281, | |
| "learning_rate": 9.470962047661077e-05, | |
| "loss": 3.4181, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.7701103687286377, | |
| "learning_rate": 9.441541629891145e-05, | |
| "loss": 3.4121, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.784576952457428, | |
| "learning_rate": 9.412180052956752e-05, | |
| "loss": 3.4052, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 0.7364141941070557, | |
| "learning_rate": 9.382759635186819e-05, | |
| "loss": 3.4, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.7345441579818726, | |
| "learning_rate": 9.353339217416887e-05, | |
| "loss": 3.3942, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.7558589577674866, | |
| "learning_rate": 9.323918799646955e-05, | |
| "loss": 3.3868, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.3813197998051418, | |
| "eval_loss": 3.5862817764282227, | |
| "eval_runtime": 154.7762, | |
| "eval_samples_per_second": 374.224, | |
| "eval_steps_per_second": 5.854, | |
| "step": 55785 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 0.773993194103241, | |
| "learning_rate": 9.294498381877022e-05, | |
| "loss": 3.3755, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.7927491068840027, | |
| "learning_rate": 9.26510738452486e-05, | |
| "loss": 3.331, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.7827730178833008, | |
| "learning_rate": 9.235686966754929e-05, | |
| "loss": 3.3343, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.7472727298736572, | |
| "learning_rate": 9.206295969402766e-05, | |
| "loss": 3.3278, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.7811399102210999, | |
| "learning_rate": 9.176875551632832e-05, | |
| "loss": 3.3254, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 0.766606867313385, | |
| "learning_rate": 9.147484554280671e-05, | |
| "loss": 3.3241, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.7435698509216309, | |
| "learning_rate": 9.118064136510739e-05, | |
| "loss": 3.3181, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.7653093338012695, | |
| "learning_rate": 9.088643718740807e-05, | |
| "loss": 3.3192, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.7305266261100769, | |
| "learning_rate": 9.059252721388644e-05, | |
| "loss": 3.3183, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.7659902572631836, | |
| "learning_rate": 9.029832303618711e-05, | |
| "loss": 3.3157, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.7410975098609924, | |
| "learning_rate": 9.000411885848779e-05, | |
| "loss": 3.3097, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.7442382574081421, | |
| "learning_rate": 8.970991468078847e-05, | |
| "loss": 3.3061, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 0.7370105981826782, | |
| "learning_rate": 8.941600470726684e-05, | |
| "loss": 3.3009, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.7415352463722229, | |
| "learning_rate": 8.912180052956752e-05, | |
| "loss": 3.3004, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.7600846886634827, | |
| "learning_rate": 8.88275963518682e-05, | |
| "loss": 3.2995, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 0.7356410622596741, | |
| "learning_rate": 8.853368637834658e-05, | |
| "loss": 3.2902, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.7762948870658875, | |
| "learning_rate": 8.823948220064724e-05, | |
| "loss": 3.2917, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 0.7024356126785278, | |
| "learning_rate": 8.794557222712563e-05, | |
| "loss": 3.2914, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.7472354173660278, | |
| "learning_rate": 8.765166225360401e-05, | |
| "loss": 3.2904, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.3888149140749002, | |
| "eval_loss": 3.5048940181732178, | |
| "eval_runtime": 154.2414, | |
| "eval_samples_per_second": 375.522, | |
| "eval_steps_per_second": 5.874, | |
| "step": 74380 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.7550144791603088, | |
| "learning_rate": 8.735745807590468e-05, | |
| "loss": 3.2496, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 0.777237057685852, | |
| "learning_rate": 8.706325389820536e-05, | |
| "loss": 3.2262, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 0.7594589591026306, | |
| "learning_rate": 8.676904972050603e-05, | |
| "loss": 3.2331, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 0.7573174238204956, | |
| "learning_rate": 8.647513974698441e-05, | |
| "loss": 3.2389, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 0.7608649134635925, | |
| "learning_rate": 8.61812297734628e-05, | |
| "loss": 3.2343, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.7639446258544922, | |
| "learning_rate": 8.588702559576346e-05, | |
| "loss": 3.2358, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 0.7428612112998962, | |
| "learning_rate": 8.559282141806415e-05, | |
| "loss": 3.2345, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 0.7416815161705017, | |
| "learning_rate": 8.529861724036483e-05, | |
| "loss": 3.2335, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 0.743262529373169, | |
| "learning_rate": 8.50044130626655e-05, | |
| "loss": 3.2331, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 0.7576801776885986, | |
| "learning_rate": 8.471020888496618e-05, | |
| "loss": 3.2308, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 0.7443795204162598, | |
| "learning_rate": 8.441659311562225e-05, | |
| "loss": 3.2345, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 0.7537975907325745, | |
| "learning_rate": 8.412238893792293e-05, | |
| "loss": 3.2282, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 0.7592223286628723, | |
| "learning_rate": 8.38284789644013e-05, | |
| "loss": 3.2234, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 0.7689176201820374, | |
| "learning_rate": 8.353427478670197e-05, | |
| "loss": 3.23, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 0.7601090669631958, | |
| "learning_rate": 8.324007060900265e-05, | |
| "loss": 3.2255, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 0.7371891736984253, | |
| "learning_rate": 8.294586643130333e-05, | |
| "loss": 3.2249, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 0.748680830001831, | |
| "learning_rate": 8.265195645778171e-05, | |
| "loss": 3.2179, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 0.7594068050384521, | |
| "learning_rate": 8.235775228008238e-05, | |
| "loss": 3.2198, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.39462156189260067, | |
| "eval_loss": 3.4777488708496094, | |
| "eval_runtime": 155.4435, | |
| "eval_samples_per_second": 372.618, | |
| "eval_steps_per_second": 5.828, | |
| "step": 92975 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.7388062477111816, | |
| "learning_rate": 8.206384230656075e-05, | |
| "loss": 3.2132, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "grad_norm": 0.7237547636032104, | |
| "learning_rate": 8.176963812886143e-05, | |
| "loss": 3.1641, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "grad_norm": 0.7498576045036316, | |
| "learning_rate": 8.14754339511621e-05, | |
| "loss": 3.1675, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 0.7394213676452637, | |
| "learning_rate": 8.118122977346278e-05, | |
| "loss": 3.1671, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "grad_norm": 0.809339165687561, | |
| "learning_rate": 8.088731979994117e-05, | |
| "loss": 3.1691, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "grad_norm": 0.7464585900306702, | |
| "learning_rate": 8.059340982641954e-05, | |
| "loss": 3.1732, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "grad_norm": 0.7435871958732605, | |
| "learning_rate": 8.02992056487202e-05, | |
| "loss": 3.1721, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 5.38, | |
| "grad_norm": 0.7047058939933777, | |
| "learning_rate": 8.000500147102089e-05, | |
| "loss": 3.1741, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 0.7210160493850708, | |
| "learning_rate": 7.971079729332157e-05, | |
| "loss": 3.1689, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "grad_norm": 0.736964225769043, | |
| "learning_rate": 7.941688731979995e-05, | |
| "loss": 3.1715, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "grad_norm": 0.7582164406776428, | |
| "learning_rate": 7.912268314210062e-05, | |
| "loss": 3.1701, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "grad_norm": 0.7246204614639282, | |
| "learning_rate": 7.882877316857899e-05, | |
| "loss": 3.1699, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "grad_norm": 0.7599069476127625, | |
| "learning_rate": 7.853456899087967e-05, | |
| "loss": 3.1718, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.773999035358429, | |
| "learning_rate": 7.824065901735806e-05, | |
| "loss": 3.1723, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 0.7431994080543518, | |
| "learning_rate": 7.794645483965872e-05, | |
| "loss": 3.1735, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "grad_norm": 0.7256425619125366, | |
| "learning_rate": 7.76525448661371e-05, | |
| "loss": 3.1692, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "grad_norm": 0.7594722509384155, | |
| "learning_rate": 7.735834068843777e-05, | |
| "loss": 3.1705, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 0.7491447925567627, | |
| "learning_rate": 7.706413651073846e-05, | |
| "loss": 3.1639, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 5.97, | |
| "grad_norm": 0.7195931077003479, | |
| "learning_rate": 7.677022653721683e-05, | |
| "loss": 3.1698, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.3970414015547945, | |
| "eval_loss": 3.4524190425872803, | |
| "eval_runtime": 154.546, | |
| "eval_samples_per_second": 374.782, | |
| "eval_steps_per_second": 5.862, | |
| "step": 111570 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 0.7680323123931885, | |
| "learning_rate": 7.647631656369521e-05, | |
| "loss": 3.1432, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "grad_norm": 0.789788544178009, | |
| "learning_rate": 7.618211238599588e-05, | |
| "loss": 3.1103, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 6.13, | |
| "grad_norm": 0.7464686632156372, | |
| "learning_rate": 7.588790820829656e-05, | |
| "loss": 3.1139, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 6.18, | |
| "grad_norm": 0.728563129901886, | |
| "learning_rate": 7.559370403059724e-05, | |
| "loss": 3.1196, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 6.24, | |
| "grad_norm": 0.7645929455757141, | |
| "learning_rate": 7.529979405707561e-05, | |
| "loss": 3.1213, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "grad_norm": 0.758517324924469, | |
| "learning_rate": 7.500558987937629e-05, | |
| "loss": 3.1222, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "grad_norm": 0.7515525817871094, | |
| "learning_rate": 7.471138570167697e-05, | |
| "loss": 3.1223, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.7567011713981628, | |
| "learning_rate": 7.441747572815534e-05, | |
| "loss": 3.1257, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 6.45, | |
| "grad_norm": 0.7746102809906006, | |
| "learning_rate": 7.412327155045603e-05, | |
| "loss": 3.1281, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 6.51, | |
| "grad_norm": 0.718424379825592, | |
| "learning_rate": 7.38293615769344e-05, | |
| "loss": 3.127, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "grad_norm": 0.7374104261398315, | |
| "learning_rate": 7.353545160341278e-05, | |
| "loss": 3.1292, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "grad_norm": 0.7483794093132019, | |
| "learning_rate": 7.324124742571345e-05, | |
| "loss": 3.1241, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 0.7430418133735657, | |
| "learning_rate": 7.294704324801413e-05, | |
| "loss": 3.1271, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "grad_norm": 0.7219746112823486, | |
| "learning_rate": 7.265283907031481e-05, | |
| "loss": 3.1297, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "grad_norm": 0.7190717458724976, | |
| "learning_rate": 7.235892909679318e-05, | |
| "loss": 3.1279, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 6.83, | |
| "grad_norm": 0.7062709331512451, | |
| "learning_rate": 7.206472491909385e-05, | |
| "loss": 3.1283, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 0.7568720579147339, | |
| "learning_rate": 7.177052074139453e-05, | |
| "loss": 3.1258, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 6.94, | |
| "grad_norm": 0.7285739779472351, | |
| "learning_rate": 7.147631656369521e-05, | |
| "loss": 3.1309, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "grad_norm": 0.732219398021698, | |
| "learning_rate": 7.11824065901736e-05, | |
| "loss": 3.1232, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.39934717198663927, | |
| "eval_loss": 3.4485130310058594, | |
| "eval_runtime": 155.2904, | |
| "eval_samples_per_second": 372.985, | |
| "eval_steps_per_second": 5.834, | |
| "step": 130165 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "grad_norm": 0.7357897162437439, | |
| "learning_rate": 7.088820241247426e-05, | |
| "loss": 3.0768, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.7629541158676147, | |
| "learning_rate": 7.059429243895263e-05, | |
| "loss": 3.0708, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "grad_norm": 0.730060875415802, | |
| "learning_rate": 7.030008826125331e-05, | |
| "loss": 3.0762, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "grad_norm": 0.7635222673416138, | |
| "learning_rate": 7.000588408355398e-05, | |
| "loss": 3.0795, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 7.26, | |
| "grad_norm": 0.7490749955177307, | |
| "learning_rate": 6.971167990585466e-05, | |
| "loss": 3.0851, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "grad_norm": 0.7679426670074463, | |
| "learning_rate": 6.941747572815534e-05, | |
| "loss": 3.0834, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "grad_norm": 0.7808982729911804, | |
| "learning_rate": 6.912385995881142e-05, | |
| "loss": 3.0839, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 7.42, | |
| "grad_norm": 0.7638462781906128, | |
| "learning_rate": 6.882965578111209e-05, | |
| "loss": 3.0879, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "grad_norm": 0.7623077630996704, | |
| "learning_rate": 6.853574580759047e-05, | |
| "loss": 3.0871, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 7.53, | |
| "grad_norm": 0.792862594127655, | |
| "learning_rate": 6.824154162989115e-05, | |
| "loss": 3.0928, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "grad_norm": 0.7692000269889832, | |
| "learning_rate": 6.794733745219183e-05, | |
| "loss": 3.0905, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "grad_norm": 0.7705720663070679, | |
| "learning_rate": 6.76537216828479e-05, | |
| "loss": 3.0887, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "grad_norm": 0.7785929441452026, | |
| "learning_rate": 6.735951750514857e-05, | |
| "loss": 3.086, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 7.74, | |
| "grad_norm": 0.7842291593551636, | |
| "learning_rate": 6.706531332744925e-05, | |
| "loss": 3.0944, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.7452487349510193, | |
| "learning_rate": 6.677110914974994e-05, | |
| "loss": 3.0946, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 7.85, | |
| "grad_norm": 0.7431057691574097, | |
| "learning_rate": 6.64771991762283e-05, | |
| "loss": 3.0915, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 7.91, | |
| "grad_norm": 0.7541413903236389, | |
| "learning_rate": 6.618299499852897e-05, | |
| "loss": 3.0904, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 7.96, | |
| "grad_norm": 0.749091386795044, | |
| "learning_rate": 6.588908502500736e-05, | |
| "loss": 3.0906, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.40129770208274423, | |
| "eval_loss": 3.431504487991333, | |
| "eval_runtime": 154.2933, | |
| "eval_samples_per_second": 375.395, | |
| "eval_steps_per_second": 5.872, | |
| "step": 148760 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 0.7683171629905701, | |
| "learning_rate": 6.559488084730804e-05, | |
| "loss": 3.0779, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "grad_norm": 0.7724857330322266, | |
| "learning_rate": 6.530067666960871e-05, | |
| "loss": 3.0379, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "grad_norm": 0.8141443729400635, | |
| "learning_rate": 6.500676669608709e-05, | |
| "loss": 3.0419, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "grad_norm": 0.7771214842796326, | |
| "learning_rate": 6.471256251838776e-05, | |
| "loss": 3.0422, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 8.23, | |
| "grad_norm": 0.7904142141342163, | |
| "learning_rate": 6.441894674904385e-05, | |
| "loss": 3.0467, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 8.28, | |
| "grad_norm": 0.7604620456695557, | |
| "learning_rate": 6.412474257134453e-05, | |
| "loss": 3.0486, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 8.34, | |
| "grad_norm": 0.7625133991241455, | |
| "learning_rate": 6.38305383936452e-05, | |
| "loss": 3.0462, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 8.39, | |
| "grad_norm": 0.7529191970825195, | |
| "learning_rate": 6.353633421594588e-05, | |
| "loss": 3.0526, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 8.44, | |
| "grad_norm": 0.7692025899887085, | |
| "learning_rate": 6.324213003824654e-05, | |
| "loss": 3.0549, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.8405176997184753, | |
| "learning_rate": 6.294851426890263e-05, | |
| "loss": 3.0587, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 8.55, | |
| "grad_norm": 0.7452796101570129, | |
| "learning_rate": 6.26543100912033e-05, | |
| "loss": 3.0599, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.7925686240196228, | |
| "learning_rate": 6.236010591350398e-05, | |
| "loss": 3.0573, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 8.66, | |
| "grad_norm": 0.7709225416183472, | |
| "learning_rate": 6.206590173580466e-05, | |
| "loss": 3.0581, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 8.71, | |
| "grad_norm": 0.7684595584869385, | |
| "learning_rate": 6.177169755810533e-05, | |
| "loss": 3.0561, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "grad_norm": 0.7561137676239014, | |
| "learning_rate": 6.14780817887614e-05, | |
| "loss": 3.0606, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 8.82, | |
| "grad_norm": 0.7704218029975891, | |
| "learning_rate": 6.118387761106208e-05, | |
| "loss": 3.0631, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 8.87, | |
| "grad_norm": 0.7560951709747314, | |
| "learning_rate": 6.088967343336276e-05, | |
| "loss": 3.0601, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "grad_norm": 0.7689745426177979, | |
| "learning_rate": 6.0595469255663425e-05, | |
| "loss": 3.0614, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 8.98, | |
| "grad_norm": 0.7952526807785034, | |
| "learning_rate": 6.0301265077964107e-05, | |
| "loss": 3.0612, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.4034028100934743, | |
| "eval_loss": 3.4061903953552246, | |
| "eval_runtime": 155.0038, | |
| "eval_samples_per_second": 373.675, | |
| "eval_steps_per_second": 5.845, | |
| "step": 167355 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "grad_norm": 0.7790576219558716, | |
| "learning_rate": 6.0007355104442484e-05, | |
| "loss": 3.0236, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "grad_norm": 0.8073156476020813, | |
| "learning_rate": 5.9713150926743165e-05, | |
| "loss": 3.0064, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 9.14, | |
| "grad_norm": 0.763963520526886, | |
| "learning_rate": 5.9419240953221536e-05, | |
| "loss": 3.0089, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.8140245676040649, | |
| "learning_rate": 5.912503677552221e-05, | |
| "loss": 3.0168, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 9.25, | |
| "grad_norm": 0.8344805240631104, | |
| "learning_rate": 5.883083259782289e-05, | |
| "loss": 3.014, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.7807629108428955, | |
| "learning_rate": 5.853692262430127e-05, | |
| "loss": 3.0188, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 9.36, | |
| "grad_norm": 0.7882916331291199, | |
| "learning_rate": 5.824271844660194e-05, | |
| "loss": 3.0272, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 9.41, | |
| "grad_norm": 0.8071965575218201, | |
| "learning_rate": 5.794880847308032e-05, | |
| "loss": 3.0258, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 9.46, | |
| "grad_norm": 0.764521598815918, | |
| "learning_rate": 5.7654604295381e-05, | |
| "loss": 3.0297, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 9.52, | |
| "grad_norm": 0.7887580990791321, | |
| "learning_rate": 5.736069432185938e-05, | |
| "loss": 3.0312, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "grad_norm": 0.8036340475082397, | |
| "learning_rate": 5.7066490144160047e-05, | |
| "loss": 3.0273, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 9.63, | |
| "grad_norm": 0.775082528591156, | |
| "learning_rate": 5.677228596646073e-05, | |
| "loss": 3.0251, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 9.68, | |
| "grad_norm": 0.8117908239364624, | |
| "learning_rate": 5.6478375992939105e-05, | |
| "loss": 3.0285, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 9.73, | |
| "grad_norm": 0.7747148871421814, | |
| "learning_rate": 5.6184171815239786e-05, | |
| "loss": 3.0319, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 9.79, | |
| "grad_norm": 0.7782384753227234, | |
| "learning_rate": 5.5889967637540454e-05, | |
| "loss": 3.0362, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 9.84, | |
| "grad_norm": 0.7501439452171326, | |
| "learning_rate": 5.559605766401883e-05, | |
| "loss": 3.0366, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.8104560375213623, | |
| "learning_rate": 5.530214769049721e-05, | |
| "loss": 3.0331, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "grad_norm": 0.7779791951179504, | |
| "learning_rate": 5.500794351279789e-05, | |
| "loss": 3.0318, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.40432080540131915, | |
| "eval_loss": 3.4155848026275635, | |
| "eval_runtime": 154.5051, | |
| "eval_samples_per_second": 374.881, | |
| "eval_steps_per_second": 5.864, | |
| "step": 185950 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.7693597674369812, | |
| "learning_rate": 5.471373933509856e-05, | |
| "loss": 3.0342, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "grad_norm": 0.7823147773742676, | |
| "learning_rate": 5.441953515739924e-05, | |
| "loss": 2.9808, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 10.11, | |
| "grad_norm": 0.8073381781578064, | |
| "learning_rate": 5.4125625183877616e-05, | |
| "loss": 2.9841, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 10.16, | |
| "grad_norm": 0.8028452396392822, | |
| "learning_rate": 5.38314210061783e-05, | |
| "loss": 2.991, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 10.22, | |
| "grad_norm": 0.8209859132766724, | |
| "learning_rate": 5.3537216828478965e-05, | |
| "loss": 2.9888, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 10.27, | |
| "grad_norm": 0.8133888244628906, | |
| "learning_rate": 5.324330685495734e-05, | |
| "loss": 2.9924, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 10.33, | |
| "grad_norm": 0.8061344623565674, | |
| "learning_rate": 5.294939688143572e-05, | |
| "loss": 2.9986, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 10.38, | |
| "grad_norm": 0.8320294618606567, | |
| "learning_rate": 5.26551927037364e-05, | |
| "loss": 2.9994, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "grad_norm": 0.7931802868843079, | |
| "learning_rate": 5.236098852603707e-05, | |
| "loss": 3.0003, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 10.49, | |
| "grad_norm": 0.8186710476875305, | |
| "learning_rate": 5.206678434833775e-05, | |
| "loss": 2.9976, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 10.54, | |
| "grad_norm": 0.7832393646240234, | |
| "learning_rate": 5.177258017063843e-05, | |
| "loss": 3.0025, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 10.59, | |
| "grad_norm": 0.7945308089256287, | |
| "learning_rate": 5.1478670197116795e-05, | |
| "loss": 3.0058, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 10.65, | |
| "grad_norm": 0.7938826084136963, | |
| "learning_rate": 5.1184466019417476e-05, | |
| "loss": 3.0027, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.8209722638130188, | |
| "learning_rate": 5.089055604589585e-05, | |
| "loss": 3.0018, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 10.76, | |
| "grad_norm": 0.7734892964363098, | |
| "learning_rate": 5.0596351868196535e-05, | |
| "loss": 3.0049, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 10.81, | |
| "grad_norm": 0.7825707793235779, | |
| "learning_rate": 5.03021476904972e-05, | |
| "loss": 3.005, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 10.86, | |
| "grad_norm": 0.7856258153915405, | |
| "learning_rate": 5.000823771697558e-05, | |
| "loss": 3.0135, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 10.92, | |
| "grad_norm": 0.7734152674674988, | |
| "learning_rate": 4.971403353927626e-05, | |
| "loss": 3.014, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 10.97, | |
| "grad_norm": 0.8184096813201904, | |
| "learning_rate": 4.9419829361576935e-05, | |
| "loss": 3.0102, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.404269548144101, | |
| "eval_loss": 3.4182839393615723, | |
| "eval_runtime": 154.7955, | |
| "eval_samples_per_second": 374.177, | |
| "eval_steps_per_second": 5.853, | |
| "step": 204545 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 0.8115602135658264, | |
| "learning_rate": 4.912591938805531e-05, | |
| "loss": 2.9826, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 11.08, | |
| "grad_norm": 0.8254956007003784, | |
| "learning_rate": 4.883171521035599e-05, | |
| "loss": 2.9594, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 11.13, | |
| "grad_norm": 0.8089250922203064, | |
| "learning_rate": 4.8537805236834364e-05, | |
| "loss": 2.9623, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 11.19, | |
| "grad_norm": 0.8682363629341125, | |
| "learning_rate": 4.824360105913504e-05, | |
| "loss": 2.9703, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 11.24, | |
| "grad_norm": 0.8310881853103638, | |
| "learning_rate": 4.794939688143572e-05, | |
| "loss": 2.9689, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 11.29, | |
| "grad_norm": 0.8337730765342712, | |
| "learning_rate": 4.7655781112091793e-05, | |
| "loss": 2.9722, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 11.35, | |
| "grad_norm": 0.800062358379364, | |
| "learning_rate": 4.736157693439247e-05, | |
| "loss": 2.9751, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.8332222700119019, | |
| "learning_rate": 4.706737275669314e-05, | |
| "loss": 2.978, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 11.45, | |
| "grad_norm": 0.8201690912246704, | |
| "learning_rate": 4.6773168578993824e-05, | |
| "loss": 2.9735, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 11.51, | |
| "grad_norm": 0.7885575890541077, | |
| "learning_rate": 4.64792586054722e-05, | |
| "loss": 2.9767, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 11.56, | |
| "grad_norm": 0.823491096496582, | |
| "learning_rate": 4.618534863195058e-05, | |
| "loss": 2.9749, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 11.62, | |
| "grad_norm": 0.8457081317901611, | |
| "learning_rate": 4.589114445425125e-05, | |
| "loss": 2.9846, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 11.67, | |
| "grad_norm": 0.8291323781013489, | |
| "learning_rate": 4.559694027655193e-05, | |
| "loss": 2.9825, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 11.72, | |
| "grad_norm": 0.7864636182785034, | |
| "learning_rate": 4.530273609885261e-05, | |
| "loss": 2.9797, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 11.78, | |
| "grad_norm": 0.8072969913482666, | |
| "learning_rate": 4.500853192115328e-05, | |
| "loss": 2.9839, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 11.83, | |
| "grad_norm": 0.8188951015472412, | |
| "learning_rate": 4.471462194763166e-05, | |
| "loss": 2.9857, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 11.88, | |
| "grad_norm": 0.8006284832954407, | |
| "learning_rate": 4.4420417769932335e-05, | |
| "loss": 2.9883, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 11.94, | |
| "grad_norm": 0.831984281539917, | |
| "learning_rate": 4.412650779641071e-05, | |
| "loss": 2.9889, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 11.99, | |
| "grad_norm": 0.8254017233848572, | |
| "learning_rate": 4.383230361871139e-05, | |
| "loss": 2.9841, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.40522657420744224, | |
| "eval_loss": 3.4149065017700195, | |
| "eval_runtime": 154.3611, | |
| "eval_samples_per_second": 375.231, | |
| "eval_steps_per_second": 5.869, | |
| "step": 223140 | |
| }, | |
| { | |
| "epoch": 12.05, | |
| "grad_norm": 0.8266258239746094, | |
| "learning_rate": 4.353809944101207e-05, | |
| "loss": 2.9382, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.8763071894645691, | |
| "learning_rate": 4.3244189467490445e-05, | |
| "loss": 2.9438, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 12.15, | |
| "grad_norm": 0.8461934924125671, | |
| "learning_rate": 4.294998528979112e-05, | |
| "loss": 2.9482, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 12.21, | |
| "grad_norm": 0.8333559036254883, | |
| "learning_rate": 4.2655781112091794e-05, | |
| "loss": 2.9442, | |
| "step": 227000 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "grad_norm": 0.8699563145637512, | |
| "learning_rate": 4.2361576934392475e-05, | |
| "loss": 2.9488, | |
| "step": 228000 | |
| }, | |
| { | |
| "epoch": 12.32, | |
| "grad_norm": 0.8646968603134155, | |
| "learning_rate": 4.2068255369226244e-05, | |
| "loss": 2.9477, | |
| "step": 229000 | |
| }, | |
| { | |
| "epoch": 12.37, | |
| "grad_norm": 0.819316565990448, | |
| "learning_rate": 4.177405119152692e-05, | |
| "loss": 2.9492, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 12.42, | |
| "grad_norm": 0.8070812225341797, | |
| "learning_rate": 4.14798470138276e-05, | |
| "loss": 2.9593, | |
| "step": 231000 | |
| }, | |
| { | |
| "epoch": 12.48, | |
| "grad_norm": 0.8208472728729248, | |
| "learning_rate": 4.1185642836128275e-05, | |
| "loss": 2.95, | |
| "step": 232000 | |
| }, | |
| { | |
| "epoch": 12.53, | |
| "grad_norm": 0.8398792147636414, | |
| "learning_rate": 4.089173286260665e-05, | |
| "loss": 2.9551, | |
| "step": 233000 | |
| }, | |
| { | |
| "epoch": 12.58, | |
| "grad_norm": 0.8196675181388855, | |
| "learning_rate": 4.0597528684907326e-05, | |
| "loss": 2.9601, | |
| "step": 234000 | |
| }, | |
| { | |
| "epoch": 12.64, | |
| "grad_norm": 0.8362085223197937, | |
| "learning_rate": 4.030332450720801e-05, | |
| "loss": 2.9624, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 12.69, | |
| "grad_norm": 0.8041438460350037, | |
| "learning_rate": 4.000912032950868e-05, | |
| "loss": 2.9616, | |
| "step": 236000 | |
| }, | |
| { | |
| "epoch": 12.75, | |
| "grad_norm": 0.828012228012085, | |
| "learning_rate": 3.971521035598706e-05, | |
| "loss": 2.9653, | |
| "step": 237000 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.8429494500160217, | |
| "learning_rate": 3.9421006178287734e-05, | |
| "loss": 2.9645, | |
| "step": 238000 | |
| }, | |
| { | |
| "epoch": 12.85, | |
| "grad_norm": 0.8595982789993286, | |
| "learning_rate": 3.912680200058841e-05, | |
| "loss": 2.9622, | |
| "step": 239000 | |
| }, | |
| { | |
| "epoch": 12.91, | |
| "grad_norm": 0.8136680722236633, | |
| "learning_rate": 3.883259782288909e-05, | |
| "loss": 2.9669, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 12.96, | |
| "grad_norm": 0.8243408203125, | |
| "learning_rate": 3.853898205354516e-05, | |
| "loss": 2.9673, | |
| "step": 241000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.40606415675396323, | |
| "eval_loss": 3.4119691848754883, | |
| "eval_runtime": 154.3522, | |
| "eval_samples_per_second": 375.252, | |
| "eval_steps_per_second": 5.87, | |
| "step": 241735 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 0.8334729671478271, | |
| "learning_rate": 3.824477787584584e-05, | |
| "loss": 2.9532, | |
| "step": 242000 | |
| }, | |
| { | |
| "epoch": 13.07, | |
| "grad_norm": 0.8407880663871765, | |
| "learning_rate": 3.795057369814651e-05, | |
| "loss": 2.9188, | |
| "step": 243000 | |
| }, | |
| { | |
| "epoch": 13.12, | |
| "grad_norm": 0.8726187944412231, | |
| "learning_rate": 3.765666372462489e-05, | |
| "loss": 2.9233, | |
| "step": 244000 | |
| }, | |
| { | |
| "epoch": 13.18, | |
| "grad_norm": 0.8489488959312439, | |
| "learning_rate": 3.7362459546925564e-05, | |
| "loss": 2.9207, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 13.23, | |
| "grad_norm": 0.8428245186805725, | |
| "learning_rate": 3.706854957340394e-05, | |
| "loss": 2.9223, | |
| "step": 246000 | |
| }, | |
| { | |
| "epoch": 13.28, | |
| "grad_norm": 0.872539222240448, | |
| "learning_rate": 3.6774345395704615e-05, | |
| "loss": 2.9313, | |
| "step": 247000 | |
| }, | |
| { | |
| "epoch": 13.34, | |
| "grad_norm": 0.8671319484710693, | |
| "learning_rate": 3.648043542218299e-05, | |
| "loss": 2.9347, | |
| "step": 248000 | |
| }, | |
| { | |
| "epoch": 13.39, | |
| "grad_norm": 0.8495607972145081, | |
| "learning_rate": 3.6186231244483674e-05, | |
| "loss": 2.9335, | |
| "step": 249000 | |
| }, | |
| { | |
| "epoch": 13.44, | |
| "grad_norm": 0.8542978763580322, | |
| "learning_rate": 3.589232127096205e-05, | |
| "loss": 2.9346, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 13.5, | |
| "grad_norm": 0.8730102777481079, | |
| "learning_rate": 3.5598117093262726e-05, | |
| "loss": 2.9363, | |
| "step": 251000 | |
| }, | |
| { | |
| "epoch": 13.55, | |
| "grad_norm": 0.8560141921043396, | |
| "learning_rate": 3.53039129155634e-05, | |
| "loss": 2.9399, | |
| "step": 252000 | |
| }, | |
| { | |
| "epoch": 13.61, | |
| "grad_norm": 0.8690813779830933, | |
| "learning_rate": 3.5009708737864075e-05, | |
| "loss": 2.9368, | |
| "step": 253000 | |
| }, | |
| { | |
| "epoch": 13.66, | |
| "grad_norm": 0.8386597633361816, | |
| "learning_rate": 3.471579876434246e-05, | |
| "loss": 2.9394, | |
| "step": 254000 | |
| }, | |
| { | |
| "epoch": 13.71, | |
| "grad_norm": 0.8756780028343201, | |
| "learning_rate": 3.4421888790820836e-05, | |
| "loss": 2.9412, | |
| "step": 255000 | |
| }, | |
| { | |
| "epoch": 13.77, | |
| "grad_norm": 0.8863993287086487, | |
| "learning_rate": 3.412768461312151e-05, | |
| "loss": 2.9428, | |
| "step": 256000 | |
| }, | |
| { | |
| "epoch": 13.82, | |
| "grad_norm": 0.8482047319412231, | |
| "learning_rate": 3.3833480435422185e-05, | |
| "loss": 2.9459, | |
| "step": 257000 | |
| }, | |
| { | |
| "epoch": 13.87, | |
| "grad_norm": 0.8722236156463623, | |
| "learning_rate": 3.353957046190056e-05, | |
| "loss": 2.9462, | |
| "step": 258000 | |
| }, | |
| { | |
| "epoch": 13.93, | |
| "grad_norm": 0.8315401077270508, | |
| "learning_rate": 3.324536628420124e-05, | |
| "loss": 2.9471, | |
| "step": 259000 | |
| }, | |
| { | |
| "epoch": 13.98, | |
| "grad_norm": 0.8811691403388977, | |
| "learning_rate": 3.295116210650192e-05, | |
| "loss": 2.9473, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.40687520376103314, | |
| "eval_loss": 3.4053232669830322, | |
| "eval_runtime": 154.4687, | |
| "eval_samples_per_second": 374.969, | |
| "eval_steps_per_second": 5.865, | |
| "step": 260330 | |
| }, | |
| { | |
| "epoch": 14.04, | |
| "grad_norm": 0.8860201239585876, | |
| "learning_rate": 3.265695792880259e-05, | |
| "loss": 2.9136, | |
| "step": 261000 | |
| }, | |
| { | |
| "epoch": 14.09, | |
| "grad_norm": 0.8469398021697998, | |
| "learning_rate": 3.236275375110327e-05, | |
| "loss": 2.9043, | |
| "step": 262000 | |
| }, | |
| { | |
| "epoch": 14.14, | |
| "grad_norm": 0.9017032384872437, | |
| "learning_rate": 3.2068843777581644e-05, | |
| "loss": 2.9092, | |
| "step": 263000 | |
| }, | |
| { | |
| "epoch": 14.2, | |
| "grad_norm": 0.9053052663803101, | |
| "learning_rate": 3.1774639599882325e-05, | |
| "loss": 2.9074, | |
| "step": 264000 | |
| }, | |
| { | |
| "epoch": 14.25, | |
| "grad_norm": 0.8582386374473572, | |
| "learning_rate": 3.1480729626360696e-05, | |
| "loss": 2.9118, | |
| "step": 265000 | |
| }, | |
| { | |
| "epoch": 14.3, | |
| "grad_norm": 0.8556872010231018, | |
| "learning_rate": 3.118652544866138e-05, | |
| "loss": 2.9124, | |
| "step": 266000 | |
| }, | |
| { | |
| "epoch": 14.36, | |
| "grad_norm": 0.8774760365486145, | |
| "learning_rate": 3.089261547513975e-05, | |
| "loss": 2.9127, | |
| "step": 267000 | |
| }, | |
| { | |
| "epoch": 14.41, | |
| "grad_norm": 0.8823800086975098, | |
| "learning_rate": 3.059841129744043e-05, | |
| "loss": 2.9129, | |
| "step": 268000 | |
| }, | |
| { | |
| "epoch": 14.47, | |
| "grad_norm": 0.8662067651748657, | |
| "learning_rate": 3.03042071197411e-05, | |
| "loss": 2.9197, | |
| "step": 269000 | |
| }, | |
| { | |
| "epoch": 14.52, | |
| "grad_norm": 0.8474991321563721, | |
| "learning_rate": 3.0010002942041775e-05, | |
| "loss": 2.9196, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 14.57, | |
| "grad_norm": 0.8806105256080627, | |
| "learning_rate": 2.9716092968520155e-05, | |
| "loss": 2.922, | |
| "step": 271000 | |
| }, | |
| { | |
| "epoch": 14.63, | |
| "grad_norm": 0.8554796576499939, | |
| "learning_rate": 2.942188879082083e-05, | |
| "loss": 2.917, | |
| "step": 272000 | |
| }, | |
| { | |
| "epoch": 14.68, | |
| "grad_norm": 0.8699655532836914, | |
| "learning_rate": 2.9127684613121508e-05, | |
| "loss": 2.9254, | |
| "step": 273000 | |
| }, | |
| { | |
| "epoch": 14.74, | |
| "grad_norm": 0.8823077082633972, | |
| "learning_rate": 2.8833480435422182e-05, | |
| "loss": 2.9236, | |
| "step": 274000 | |
| }, | |
| { | |
| "epoch": 14.79, | |
| "grad_norm": 0.8708270788192749, | |
| "learning_rate": 2.8539570461900563e-05, | |
| "loss": 2.9238, | |
| "step": 275000 | |
| }, | |
| { | |
| "epoch": 14.84, | |
| "grad_norm": 0.866905152797699, | |
| "learning_rate": 2.8245366284201237e-05, | |
| "loss": 2.9211, | |
| "step": 276000 | |
| }, | |
| { | |
| "epoch": 14.9, | |
| "grad_norm": 0.9085943698883057, | |
| "learning_rate": 2.7951456310679614e-05, | |
| "loss": 2.9226, | |
| "step": 277000 | |
| }, | |
| { | |
| "epoch": 14.95, | |
| "grad_norm": 0.8788062334060669, | |
| "learning_rate": 2.7657546337157992e-05, | |
| "loss": 2.9271, | |
| "step": 278000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.4066371228703634, | |
| "eval_loss": 3.414015531539917, | |
| "eval_runtime": 154.5006, | |
| "eval_samples_per_second": 374.892, | |
| "eval_steps_per_second": 5.864, | |
| "step": 278925 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.8724709749221802, | |
| "learning_rate": 2.7363342159458666e-05, | |
| "loss": 2.9244, | |
| "step": 279000 | |
| }, | |
| { | |
| "epoch": 15.06, | |
| "grad_norm": 0.8519229292869568, | |
| "learning_rate": 2.7069432185937043e-05, | |
| "loss": 2.8883, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 15.11, | |
| "grad_norm": 0.8671460747718811, | |
| "learning_rate": 2.6775228008237718e-05, | |
| "loss": 2.8849, | |
| "step": 281000 | |
| }, | |
| { | |
| "epoch": 15.17, | |
| "grad_norm": 0.8618038892745972, | |
| "learning_rate": 2.6481318034716095e-05, | |
| "loss": 2.8883, | |
| "step": 282000 | |
| }, | |
| { | |
| "epoch": 15.22, | |
| "grad_norm": 0.8848353028297424, | |
| "learning_rate": 2.618711385701677e-05, | |
| "loss": 2.8928, | |
| "step": 283000 | |
| }, | |
| { | |
| "epoch": 15.27, | |
| "grad_norm": 0.8928247094154358, | |
| "learning_rate": 2.5893203883495147e-05, | |
| "loss": 2.8936, | |
| "step": 284000 | |
| }, | |
| { | |
| "epoch": 15.33, | |
| "grad_norm": 0.9044170379638672, | |
| "learning_rate": 2.559899970579582e-05, | |
| "loss": 2.8963, | |
| "step": 285000 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "grad_norm": 0.8573793768882751, | |
| "learning_rate": 2.5305089732274202e-05, | |
| "loss": 2.8973, | |
| "step": 286000 | |
| }, | |
| { | |
| "epoch": 15.43, | |
| "grad_norm": 0.8782808780670166, | |
| "learning_rate": 2.5010885554574877e-05, | |
| "loss": 2.9009, | |
| "step": 287000 | |
| }, | |
| { | |
| "epoch": 15.49, | |
| "grad_norm": 0.891230583190918, | |
| "learning_rate": 2.4716681376875554e-05, | |
| "loss": 2.9006, | |
| "step": 288000 | |
| }, | |
| { | |
| "epoch": 15.54, | |
| "grad_norm": 0.8635398149490356, | |
| "learning_rate": 2.442277140335393e-05, | |
| "loss": 2.9065, | |
| "step": 289000 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "grad_norm": 0.8536158800125122, | |
| "learning_rate": 2.4128861429832306e-05, | |
| "loss": 2.8981, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 15.65, | |
| "grad_norm": 0.897195041179657, | |
| "learning_rate": 2.383465725213298e-05, | |
| "loss": 2.9047, | |
| "step": 291000 | |
| }, | |
| { | |
| "epoch": 15.7, | |
| "grad_norm": 0.8918027281761169, | |
| "learning_rate": 2.3540453074433658e-05, | |
| "loss": 2.9027, | |
| "step": 292000 | |
| }, | |
| { | |
| "epoch": 15.76, | |
| "grad_norm": 0.8919987082481384, | |
| "learning_rate": 2.3246248896734336e-05, | |
| "loss": 2.9065, | |
| "step": 293000 | |
| }, | |
| { | |
| "epoch": 15.81, | |
| "grad_norm": 0.9096593260765076, | |
| "learning_rate": 2.2952044719035014e-05, | |
| "loss": 2.9096, | |
| "step": 294000 | |
| }, | |
| { | |
| "epoch": 15.86, | |
| "grad_norm": 0.9001402854919434, | |
| "learning_rate": 2.2657840541335688e-05, | |
| "loss": 2.9032, | |
| "step": 295000 | |
| }, | |
| { | |
| "epoch": 15.92, | |
| "grad_norm": 0.892467737197876, | |
| "learning_rate": 2.2363930567814065e-05, | |
| "loss": 2.9072, | |
| "step": 296000 | |
| }, | |
| { | |
| "epoch": 15.97, | |
| "grad_norm": 0.894195556640625, | |
| "learning_rate": 2.207002059429244e-05, | |
| "loss": 2.9087, | |
| "step": 297000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.40667239162533003, | |
| "eval_loss": 3.4227283000946045, | |
| "eval_runtime": 154.7066, | |
| "eval_samples_per_second": 374.392, | |
| "eval_steps_per_second": 5.856, | |
| "step": 297520 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "grad_norm": 0.8851794004440308, | |
| "learning_rate": 2.1775816416593117e-05, | |
| "loss": 2.8905, | |
| "step": 298000 | |
| }, | |
| { | |
| "epoch": 16.08, | |
| "grad_norm": 0.8866509199142456, | |
| "learning_rate": 2.1481612238893795e-05, | |
| "loss": 2.8718, | |
| "step": 299000 | |
| }, | |
| { | |
| "epoch": 16.13, | |
| "grad_norm": 0.8785372376441956, | |
| "learning_rate": 2.118799646954987e-05, | |
| "loss": 2.876, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 16.19, | |
| "grad_norm": 0.8921369910240173, | |
| "learning_rate": 2.0893792291850546e-05, | |
| "loss": 2.879, | |
| "step": 301000 | |
| }, | |
| { | |
| "epoch": 16.24, | |
| "grad_norm": 0.894282877445221, | |
| "learning_rate": 2.059958811415122e-05, | |
| "loss": 2.8786, | |
| "step": 302000 | |
| }, | |
| { | |
| "epoch": 16.29, | |
| "grad_norm": 0.9055682420730591, | |
| "learning_rate": 2.03053839364519e-05, | |
| "loss": 2.8772, | |
| "step": 303000 | |
| }, | |
| { | |
| "epoch": 16.35, | |
| "grad_norm": 0.8979899287223816, | |
| "learning_rate": 2.0011179758752573e-05, | |
| "loss": 2.8826, | |
| "step": 304000 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "grad_norm": 0.8827613592147827, | |
| "learning_rate": 1.971726978523095e-05, | |
| "loss": 2.8867, | |
| "step": 305000 | |
| }, | |
| { | |
| "epoch": 16.46, | |
| "grad_norm": 0.9219442009925842, | |
| "learning_rate": 1.9423065607531628e-05, | |
| "loss": 2.8839, | |
| "step": 306000 | |
| }, | |
| { | |
| "epoch": 16.51, | |
| "grad_norm": 0.8988754153251648, | |
| "learning_rate": 1.9128861429832303e-05, | |
| "loss": 2.8813, | |
| "step": 307000 | |
| }, | |
| { | |
| "epoch": 16.56, | |
| "grad_norm": 0.8807535767555237, | |
| "learning_rate": 1.883495145631068e-05, | |
| "loss": 2.8821, | |
| "step": 308000 | |
| }, | |
| { | |
| "epoch": 16.62, | |
| "grad_norm": 0.9281722903251648, | |
| "learning_rate": 1.8540747278611358e-05, | |
| "loss": 2.8845, | |
| "step": 309000 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "grad_norm": 0.8842557668685913, | |
| "learning_rate": 1.824713150926743e-05, | |
| "loss": 2.8861, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 16.72, | |
| "grad_norm": 0.9147974252700806, | |
| "learning_rate": 1.795292733156811e-05, | |
| "loss": 2.8917, | |
| "step": 311000 | |
| }, | |
| { | |
| "epoch": 16.78, | |
| "grad_norm": 0.9400495290756226, | |
| "learning_rate": 1.7658723153868787e-05, | |
| "loss": 2.8896, | |
| "step": 312000 | |
| }, | |
| { | |
| "epoch": 16.83, | |
| "grad_norm": 0.908082902431488, | |
| "learning_rate": 1.7364518976169465e-05, | |
| "loss": 2.8896, | |
| "step": 313000 | |
| }, | |
| { | |
| "epoch": 16.89, | |
| "grad_norm": 0.9106447696685791, | |
| "learning_rate": 1.707031479847014e-05, | |
| "loss": 2.8892, | |
| "step": 314000 | |
| }, | |
| { | |
| "epoch": 16.94, | |
| "grad_norm": 0.8716493844985962, | |
| "learning_rate": 1.6776110620770817e-05, | |
| "loss": 2.8873, | |
| "step": 315000 | |
| }, | |
| { | |
| "epoch": 16.99, | |
| "grad_norm": 0.8908538818359375, | |
| "learning_rate": 1.648220064724919e-05, | |
| "loss": 2.8915, | |
| "step": 316000 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.40704959935702034, | |
| "eval_loss": 3.4177122116088867, | |
| "eval_runtime": 154.3312, | |
| "eval_samples_per_second": 375.303, | |
| "eval_steps_per_second": 5.87, | |
| "step": 316115 | |
| }, | |
| { | |
| "epoch": 17.05, | |
| "grad_norm": 0.9286654591560364, | |
| "learning_rate": 1.6188290673727568e-05, | |
| "loss": 2.8621, | |
| "step": 317000 | |
| }, | |
| { | |
| "epoch": 17.1, | |
| "grad_norm": 0.9319996237754822, | |
| "learning_rate": 1.5894086496028243e-05, | |
| "loss": 2.86, | |
| "step": 318000 | |
| }, | |
| { | |
| "epoch": 17.16, | |
| "grad_norm": 0.941879391670227, | |
| "learning_rate": 1.560017652250662e-05, | |
| "loss": 2.8593, | |
| "step": 319000 | |
| }, | |
| { | |
| "epoch": 17.21, | |
| "grad_norm": 0.9331826567649841, | |
| "learning_rate": 1.5305972344807298e-05, | |
| "loss": 2.8631, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 17.26, | |
| "grad_norm": 0.9396593570709229, | |
| "learning_rate": 1.5011768167107972e-05, | |
| "loss": 2.8632, | |
| "step": 321000 | |
| }, | |
| { | |
| "epoch": 17.32, | |
| "grad_norm": 0.9413577914237976, | |
| "learning_rate": 1.471756398940865e-05, | |
| "loss": 2.8667, | |
| "step": 322000 | |
| }, | |
| { | |
| "epoch": 17.37, | |
| "grad_norm": 0.9180823564529419, | |
| "learning_rate": 1.4423654015887026e-05, | |
| "loss": 2.8645, | |
| "step": 323000 | |
| }, | |
| { | |
| "epoch": 17.42, | |
| "grad_norm": 0.946160614490509, | |
| "learning_rate": 1.4129744042365403e-05, | |
| "loss": 2.8681, | |
| "step": 324000 | |
| }, | |
| { | |
| "epoch": 17.48, | |
| "grad_norm": 0.9351367950439453, | |
| "learning_rate": 1.3835539864666078e-05, | |
| "loss": 2.869, | |
| "step": 325000 | |
| }, | |
| { | |
| "epoch": 17.53, | |
| "grad_norm": 0.936220109462738, | |
| "learning_rate": 1.3541335686966755e-05, | |
| "loss": 2.8684, | |
| "step": 326000 | |
| }, | |
| { | |
| "epoch": 17.59, | |
| "grad_norm": 0.9257025122642517, | |
| "learning_rate": 1.3247131509267433e-05, | |
| "loss": 2.8697, | |
| "step": 327000 | |
| }, | |
| { | |
| "epoch": 17.64, | |
| "grad_norm": 0.9127140641212463, | |
| "learning_rate": 1.2953221535745807e-05, | |
| "loss": 2.8715, | |
| "step": 328000 | |
| }, | |
| { | |
| "epoch": 17.69, | |
| "grad_norm": 0.9492712616920471, | |
| "learning_rate": 1.2659017358046485e-05, | |
| "loss": 2.8714, | |
| "step": 329000 | |
| }, | |
| { | |
| "epoch": 17.75, | |
| "grad_norm": 0.9370437860488892, | |
| "learning_rate": 1.236510738452486e-05, | |
| "loss": 2.8722, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 17.8, | |
| "grad_norm": 0.9461432695388794, | |
| "learning_rate": 1.2071197411003236e-05, | |
| "loss": 2.8723, | |
| "step": 331000 | |
| }, | |
| { | |
| "epoch": 17.85, | |
| "grad_norm": 0.9277128577232361, | |
| "learning_rate": 1.1776993233303914e-05, | |
| "loss": 2.8736, | |
| "step": 332000 | |
| }, | |
| { | |
| "epoch": 17.91, | |
| "grad_norm": 0.9499723315238953, | |
| "learning_rate": 1.1483083259782288e-05, | |
| "loss": 2.8781, | |
| "step": 333000 | |
| }, | |
| { | |
| "epoch": 17.96, | |
| "grad_norm": 0.9101009368896484, | |
| "learning_rate": 1.1188879082082966e-05, | |
| "loss": 2.8719, | |
| "step": 334000 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.40709910997113535, | |
| "eval_loss": 3.4237568378448486, | |
| "eval_runtime": 154.9706, | |
| "eval_samples_per_second": 373.755, | |
| "eval_steps_per_second": 5.846, | |
| "step": 334710 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 0.9372355341911316, | |
| "learning_rate": 1.0894674904383642e-05, | |
| "loss": 2.8658, | |
| "step": 335000 | |
| }, | |
| { | |
| "epoch": 18.07, | |
| "grad_norm": 0.9395254850387573, | |
| "learning_rate": 1.060047072668432e-05, | |
| "loss": 2.8485, | |
| "step": 336000 | |
| }, | |
| { | |
| "epoch": 18.12, | |
| "grad_norm": 0.9409775733947754, | |
| "learning_rate": 1.0306560753162696e-05, | |
| "loss": 2.8483, | |
| "step": 337000 | |
| }, | |
| { | |
| "epoch": 18.18, | |
| "grad_norm": 0.9178088903427124, | |
| "learning_rate": 1.0012650779641071e-05, | |
| "loss": 2.8486, | |
| "step": 338000 | |
| }, | |
| { | |
| "epoch": 18.23, | |
| "grad_norm": 0.9488175511360168, | |
| "learning_rate": 9.718446601941749e-06, | |
| "loss": 2.8534, | |
| "step": 339000 | |
| }, | |
| { | |
| "epoch": 18.28, | |
| "grad_norm": 0.9234158992767334, | |
| "learning_rate": 9.424242424242425e-06, | |
| "loss": 2.8575, | |
| "step": 340000 | |
| }, | |
| { | |
| "epoch": 18.34, | |
| "grad_norm": 0.9424415230751038, | |
| "learning_rate": 9.1303324507208e-06, | |
| "loss": 2.8504, | |
| "step": 341000 | |
| }, | |
| { | |
| "epoch": 18.39, | |
| "grad_norm": 0.9509896636009216, | |
| "learning_rate": 8.836128273021477e-06, | |
| "loss": 2.8565, | |
| "step": 342000 | |
| }, | |
| { | |
| "epoch": 18.45, | |
| "grad_norm": 0.9754331707954407, | |
| "learning_rate": 8.541924095322155e-06, | |
| "loss": 2.8518, | |
| "step": 343000 | |
| }, | |
| { | |
| "epoch": 18.5, | |
| "grad_norm": 0.9644212126731873, | |
| "learning_rate": 8.247719917622831e-06, | |
| "loss": 2.8551, | |
| "step": 344000 | |
| }, | |
| { | |
| "epoch": 18.55, | |
| "grad_norm": 0.9579382538795471, | |
| "learning_rate": 7.953809944101207e-06, | |
| "loss": 2.8598, | |
| "step": 345000 | |
| }, | |
| { | |
| "epoch": 18.61, | |
| "grad_norm": 0.9656188488006592, | |
| "learning_rate": 7.659605766401884e-06, | |
| "loss": 2.8556, | |
| "step": 346000 | |
| }, | |
| { | |
| "epoch": 18.66, | |
| "grad_norm": 0.9785469770431519, | |
| "learning_rate": 7.365695792880259e-06, | |
| "loss": 2.8545, | |
| "step": 347000 | |
| }, | |
| { | |
| "epoch": 18.71, | |
| "grad_norm": 0.9273431897163391, | |
| "learning_rate": 7.0717858193586356e-06, | |
| "loss": 2.8528, | |
| "step": 348000 | |
| }, | |
| { | |
| "epoch": 18.77, | |
| "grad_norm": 0.9148189425468445, | |
| "learning_rate": 6.777581641659312e-06, | |
| "loss": 2.8607, | |
| "step": 349000 | |
| }, | |
| { | |
| "epoch": 18.82, | |
| "grad_norm": 0.9610157012939453, | |
| "learning_rate": 6.483377463959989e-06, | |
| "loss": 2.8558, | |
| "step": 350000 | |
| }, | |
| { | |
| "epoch": 18.88, | |
| "grad_norm": 0.9283749461174011, | |
| "learning_rate": 6.189173286260666e-06, | |
| "loss": 2.8575, | |
| "step": 351000 | |
| }, | |
| { | |
| "epoch": 18.93, | |
| "grad_norm": 0.9314181208610535, | |
| "learning_rate": 5.894969108561342e-06, | |
| "loss": 2.8586, | |
| "step": 352000 | |
| }, | |
| { | |
| "epoch": 18.98, | |
| "grad_norm": 0.9226950407028198, | |
| "learning_rate": 5.6010591350397175e-06, | |
| "loss": 2.8512, | |
| "step": 353000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.407070962145743, | |
| "eval_loss": 3.4331676959991455, | |
| "eval_runtime": 155.2473, | |
| "eval_samples_per_second": 373.089, | |
| "eval_steps_per_second": 5.836, | |
| "step": 353305 | |
| }, | |
| { | |
| "epoch": 19.04, | |
| "grad_norm": 0.9619930386543274, | |
| "learning_rate": 5.3068549573403945e-06, | |
| "loss": 2.8452, | |
| "step": 354000 | |
| }, | |
| { | |
| "epoch": 19.09, | |
| "grad_norm": 0.9528549313545227, | |
| "learning_rate": 5.0126507796410715e-06, | |
| "loss": 2.8397, | |
| "step": 355000 | |
| }, | |
| { | |
| "epoch": 19.14, | |
| "grad_norm": 0.9460570216178894, | |
| "learning_rate": 4.718740806119447e-06, | |
| "loss": 2.8382, | |
| "step": 356000 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "grad_norm": 0.9692838191986084, | |
| "learning_rate": 4.424536628420123e-06, | |
| "loss": 2.8444, | |
| "step": 357000 | |
| }, | |
| { | |
| "epoch": 19.25, | |
| "grad_norm": 0.9561824798583984, | |
| "learning_rate": 4.1303324507208e-06, | |
| "loss": 2.8394, | |
| "step": 358000 | |
| }, | |
| { | |
| "epoch": 19.31, | |
| "grad_norm": 0.9296232461929321, | |
| "learning_rate": 3.836422477199177e-06, | |
| "loss": 2.8444, | |
| "step": 359000 | |
| }, | |
| { | |
| "epoch": 19.36, | |
| "grad_norm": 0.9455821514129639, | |
| "learning_rate": 3.5422182994998533e-06, | |
| "loss": 2.8454, | |
| "step": 360000 | |
| }, | |
| { | |
| "epoch": 19.41, | |
| "grad_norm": 0.961200475692749, | |
| "learning_rate": 3.24801412180053e-06, | |
| "loss": 2.8405, | |
| "step": 361000 | |
| }, | |
| { | |
| "epoch": 19.47, | |
| "grad_norm": 0.961736261844635, | |
| "learning_rate": 2.9541041482789055e-06, | |
| "loss": 2.8387, | |
| "step": 362000 | |
| }, | |
| { | |
| "epoch": 19.52, | |
| "grad_norm": 0.9697670340538025, | |
| "learning_rate": 2.6598999705795825e-06, | |
| "loss": 2.8446, | |
| "step": 363000 | |
| }, | |
| { | |
| "epoch": 19.58, | |
| "grad_norm": 0.9657266139984131, | |
| "learning_rate": 2.365695792880259e-06, | |
| "loss": 2.8442, | |
| "step": 364000 | |
| }, | |
| { | |
| "epoch": 19.63, | |
| "grad_norm": 0.9578980803489685, | |
| "learning_rate": 2.0714916151809357e-06, | |
| "loss": 2.8431, | |
| "step": 365000 | |
| }, | |
| { | |
| "epoch": 19.68, | |
| "grad_norm": 0.9245219230651855, | |
| "learning_rate": 1.7775816416593115e-06, | |
| "loss": 2.8413, | |
| "step": 366000 | |
| }, | |
| { | |
| "epoch": 19.74, | |
| "grad_norm": 0.9485571384429932, | |
| "learning_rate": 1.4833774639599883e-06, | |
| "loss": 2.8404, | |
| "step": 367000 | |
| }, | |
| { | |
| "epoch": 19.79, | |
| "grad_norm": 0.9614645838737488, | |
| "learning_rate": 1.189173286260665e-06, | |
| "loss": 2.8414, | |
| "step": 368000 | |
| }, | |
| { | |
| "epoch": 19.84, | |
| "grad_norm": 0.9490430951118469, | |
| "learning_rate": 8.949691085613415e-07, | |
| "loss": 2.8403, | |
| "step": 369000 | |
| }, | |
| { | |
| "epoch": 19.9, | |
| "grad_norm": 0.9546143412590027, | |
| "learning_rate": 6.010591350397176e-07, | |
| "loss": 2.8419, | |
| "step": 370000 | |
| }, | |
| { | |
| "epoch": 19.95, | |
| "grad_norm": 0.9518195986747742, | |
| "learning_rate": 3.071491615180936e-07, | |
| "loss": 2.8466, | |
| "step": 371000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.4072716245668577, | |
| "eval_loss": 3.4357750415802, | |
| "eval_runtime": 154.8025, | |
| "eval_samples_per_second": 374.161, | |
| "eval_steps_per_second": 5.853, | |
| "step": 371900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 371900, | |
| "total_flos": 1.5669257538816e+18, | |
| "train_loss": 3.1535628946688457, | |
| "train_runtime": 81351.2344, | |
| "train_samples_per_second": 146.288, | |
| "train_steps_per_second": 4.572 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 371900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 5000, | |
| "total_flos": 1.5669257538816e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |