| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.8, |
| "eval_steps": 100, |
| "global_step": 1000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "entropy": 2.5604241602122784, |
| "epoch": 0.008, |
| "grad_norm": 360.0, |
| "learning_rate": 1.8e-07, |
| "loss": 6.7873, |
| "mean_token_accuracy": 0.21116912770085036, |
| "num_tokens": 175126.0, |
| "step": 10 |
| }, |
| { |
| "entropy": 2.595553621649742, |
| "epoch": 0.016, |
| "grad_norm": 372.0, |
| "learning_rate": 3.8e-07, |
| "loss": 6.7228, |
| "mean_token_accuracy": 0.20832345820963383, |
| "num_tokens": 340696.0, |
| "step": 20 |
| }, |
| { |
| "entropy": 2.6845590390264986, |
| "epoch": 0.024, |
| "grad_norm": 270.0, |
| "learning_rate": 5.800000000000001e-07, |
| "loss": 6.3033, |
| "mean_token_accuracy": 0.21763876075856387, |
| "num_tokens": 509066.0, |
| "step": 30 |
| }, |
| { |
| "entropy": 2.8930466674268245, |
| "epoch": 0.032, |
| "grad_norm": 302.0, |
| "learning_rate": 7.8e-07, |
| "loss": 5.6306, |
| "mean_token_accuracy": 0.22912178803235292, |
| "num_tokens": 675930.0, |
| "step": 40 |
| }, |
| { |
| "entropy": 3.081789229810238, |
| "epoch": 0.04, |
| "grad_norm": 123.0, |
| "learning_rate": 9.800000000000001e-07, |
| "loss": 4.9613, |
| "mean_token_accuracy": 0.24729052558541298, |
| "num_tokens": 841685.0, |
| "step": 50 |
| }, |
| { |
| "entropy": 3.404916675388813, |
| "epoch": 0.048, |
| "grad_norm": 66.0, |
| "learning_rate": 1.1800000000000001e-06, |
| "loss": 4.2301, |
| "mean_token_accuracy": 0.27200257908552883, |
| "num_tokens": 1007956.0, |
| "step": 60 |
| }, |
| { |
| "entropy": 3.6782597199082376, |
| "epoch": 0.056, |
| "grad_norm": 38.0, |
| "learning_rate": 1.3800000000000001e-06, |
| "loss": 3.7823, |
| "mean_token_accuracy": 0.3019493265077472, |
| "num_tokens": 1174735.0, |
| "step": 70 |
| }, |
| { |
| "entropy": 3.4533004209399225, |
| "epoch": 0.064, |
| "grad_norm": 33.75, |
| "learning_rate": 1.5800000000000001e-06, |
| "loss": 3.4895, |
| "mean_token_accuracy": 0.33877146458253266, |
| "num_tokens": 1340560.0, |
| "step": 80 |
| }, |
| { |
| "entropy": 3.139299413561821, |
| "epoch": 0.072, |
| "grad_norm": 31.625, |
| "learning_rate": 1.7800000000000001e-06, |
| "loss": 3.1861, |
| "mean_token_accuracy": 0.3694635771214962, |
| "num_tokens": 1510110.0, |
| "step": 90 |
| }, |
| { |
| "entropy": 2.9895101204514503, |
| "epoch": 0.08, |
| "grad_norm": 21.25, |
| "learning_rate": 1.98e-06, |
| "loss": 2.9954, |
| "mean_token_accuracy": 0.39561883062124253, |
| "num_tokens": 1679432.0, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_coding_entropy": 2.0758819575309753, |
| "eval_coding_loss": 2.1979613304138184, |
| "eval_coding_mean_token_accuracy": 0.502328829407692, |
| "eval_coding_num_tokens": 1679432.0, |
| "eval_coding_runtime": 79.4698, |
| "eval_coding_samples_per_second": 6.292, |
| "eval_coding_steps_per_second": 3.146, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.08, |
| "eval_physics_entropy": 2.9290144443511963, |
| "eval_physics_loss": 2.974918842315674, |
| "eval_physics_mean_token_accuracy": 0.40643512630462647, |
| "eval_physics_num_tokens": 1679432.0, |
| "eval_physics_runtime": 52.0573, |
| "eval_physics_samples_per_second": 9.605, |
| "eval_physics_steps_per_second": 4.802, |
| "step": 100 |
| }, |
| { |
| "entropy": 2.8267079159617423, |
| "epoch": 0.088, |
| "grad_norm": 21.125, |
| "learning_rate": 2.1800000000000003e-06, |
| "loss": 2.8166, |
| "mean_token_accuracy": 0.4188435094431043, |
| "num_tokens": 1848521.0, |
| "step": 110 |
| }, |
| { |
| "entropy": 2.731858339905739, |
| "epoch": 0.096, |
| "grad_norm": 19.625, |
| "learning_rate": 2.38e-06, |
| "loss": 2.7192, |
| "mean_token_accuracy": 0.43734086528420446, |
| "num_tokens": 2021570.0, |
| "step": 120 |
| }, |
| { |
| "entropy": 2.742488440871239, |
| "epoch": 0.104, |
| "grad_norm": 20.5, |
| "learning_rate": 2.5800000000000003e-06, |
| "loss": 2.7106, |
| "mean_token_accuracy": 0.4448789706453681, |
| "num_tokens": 2186993.0, |
| "step": 130 |
| }, |
| { |
| "entropy": 2.546742835640907, |
| "epoch": 0.112, |
| "grad_norm": 19.875, |
| "learning_rate": 2.7800000000000005e-06, |
| "loss": 2.5347, |
| "mean_token_accuracy": 0.46661383416503666, |
| "num_tokens": 2359297.0, |
| "step": 140 |
| }, |
| { |
| "entropy": 2.5924400255084037, |
| "epoch": 0.12, |
| "grad_norm": 21.25, |
| "learning_rate": 2.9800000000000003e-06, |
| "loss": 2.5665, |
| "mean_token_accuracy": 0.4689426450058818, |
| "num_tokens": 2522375.0, |
| "step": 150 |
| }, |
| { |
| "entropy": 2.4927076123654843, |
| "epoch": 0.128, |
| "grad_norm": 21.125, |
| "learning_rate": 3.1800000000000005e-06, |
| "loss": 2.4571, |
| "mean_token_accuracy": 0.4879589693620801, |
| "num_tokens": 2683168.0, |
| "step": 160 |
| }, |
| { |
| "entropy": 2.4049933359026907, |
| "epoch": 0.136, |
| "grad_norm": 18.5, |
| "learning_rate": 3.3800000000000007e-06, |
| "loss": 2.3829, |
| "mean_token_accuracy": 0.4920692486688495, |
| "num_tokens": 2850132.0, |
| "step": 170 |
| }, |
| { |
| "entropy": 2.3835881277918816, |
| "epoch": 0.144, |
| "grad_norm": 20.5, |
| "learning_rate": 3.58e-06, |
| "loss": 2.3245, |
| "mean_token_accuracy": 0.5039734555408358, |
| "num_tokens": 3019426.0, |
| "step": 180 |
| }, |
| { |
| "entropy": 2.3372473388910295, |
| "epoch": 0.152, |
| "grad_norm": 18.25, |
| "learning_rate": 3.7800000000000002e-06, |
| "loss": 2.3194, |
| "mean_token_accuracy": 0.5051064729690552, |
| "num_tokens": 3182640.0, |
| "step": 190 |
| }, |
| { |
| "entropy": 2.27973321005702, |
| "epoch": 0.16, |
| "grad_norm": 21.125, |
| "learning_rate": 3.980000000000001e-06, |
| "loss": 2.213, |
| "mean_token_accuracy": 0.5190408078953623, |
| "num_tokens": 3354390.0, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_coding_entropy": 1.9250439591407775, |
| "eval_coding_loss": 2.1208813190460205, |
| "eval_coding_mean_token_accuracy": 0.5191248273849487, |
| "eval_coding_num_tokens": 3354390.0, |
| "eval_coding_runtime": 79.1564, |
| "eval_coding_samples_per_second": 6.317, |
| "eval_coding_steps_per_second": 3.158, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.16, |
| "eval_physics_entropy": 2.230250364303589, |
| "eval_physics_loss": 2.271095037460327, |
| "eval_physics_mean_token_accuracy": 0.5189732195138931, |
| "eval_physics_num_tokens": 3354390.0, |
| "eval_physics_runtime": 52.0613, |
| "eval_physics_samples_per_second": 9.604, |
| "eval_physics_steps_per_second": 4.802, |
| "step": 200 |
| }, |
| { |
| "entropy": 2.265408169478178, |
| "epoch": 0.168, |
| "grad_norm": 20.5, |
| "learning_rate": 4.18e-06, |
| "loss": 2.2294, |
| "mean_token_accuracy": 0.5198547914624214, |
| "num_tokens": 3521075.0, |
| "step": 210 |
| }, |
| { |
| "entropy": 2.2567644871771337, |
| "epoch": 0.176, |
| "grad_norm": 26.375, |
| "learning_rate": 4.38e-06, |
| "loss": 2.1962, |
| "mean_token_accuracy": 0.5276498578488826, |
| "num_tokens": 3684374.0, |
| "step": 220 |
| }, |
| { |
| "entropy": 2.2172055408358573, |
| "epoch": 0.184, |
| "grad_norm": 18.5, |
| "learning_rate": 4.58e-06, |
| "loss": 2.1651, |
| "mean_token_accuracy": 0.5290316773578525, |
| "num_tokens": 3848536.0, |
| "step": 230 |
| }, |
| { |
| "entropy": 2.152971290051937, |
| "epoch": 0.192, |
| "grad_norm": 18.75, |
| "learning_rate": 4.78e-06, |
| "loss": 2.1125, |
| "mean_token_accuracy": 0.5370275607332587, |
| "num_tokens": 4023915.0, |
| "step": 240 |
| }, |
| { |
| "entropy": 2.148153053969145, |
| "epoch": 0.2, |
| "grad_norm": 20.125, |
| "learning_rate": 4.980000000000001e-06, |
| "loss": 2.1104, |
| "mean_token_accuracy": 0.541122961603105, |
| "num_tokens": 4190485.0, |
| "step": 250 |
| }, |
| { |
| "entropy": 2.0243859082460403, |
| "epoch": 0.208, |
| "grad_norm": 18.125, |
| "learning_rate": 5.18e-06, |
| "loss": 1.975, |
| "mean_token_accuracy": 0.558369823731482, |
| "num_tokens": 4352045.0, |
| "step": 260 |
| }, |
| { |
| "entropy": 2.0731096126139166, |
| "epoch": 0.216, |
| "grad_norm": 19.625, |
| "learning_rate": 5.380000000000001e-06, |
| "loss": 2.0131, |
| "mean_token_accuracy": 0.5515947749838233, |
| "num_tokens": 4514042.0, |
| "step": 270 |
| }, |
| { |
| "entropy": 2.0090589359402657, |
| "epoch": 0.224, |
| "grad_norm": 17.375, |
| "learning_rate": 5.580000000000001e-06, |
| "loss": 1.9505, |
| "mean_token_accuracy": 0.5629682159051299, |
| "num_tokens": 4681371.0, |
| "step": 280 |
| }, |
| { |
| "entropy": 2.002196607738733, |
| "epoch": 0.232, |
| "grad_norm": 18.625, |
| "learning_rate": 5.78e-06, |
| "loss": 1.958, |
| "mean_token_accuracy": 0.5620508158579469, |
| "num_tokens": 4847650.0, |
| "step": 290 |
| }, |
| { |
| "entropy": 1.9909410774707794, |
| "epoch": 0.24, |
| "grad_norm": 18.375, |
| "learning_rate": 5.98e-06, |
| "loss": 1.9375, |
| "mean_token_accuracy": 0.5657549701631069, |
| "num_tokens": 5021784.0, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_coding_entropy": 1.916715157032013, |
| "eval_coding_loss": 2.0919320583343506, |
| "eval_coding_mean_token_accuracy": 0.5256761227846145, |
| "eval_coding_num_tokens": 5021784.0, |
| "eval_coding_runtime": 79.1652, |
| "eval_coding_samples_per_second": 6.316, |
| "eval_coding_steps_per_second": 3.158, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.24, |
| "eval_physics_entropy": 1.9569331226348876, |
| "eval_physics_loss": 1.9659218788146973, |
| "eval_physics_mean_token_accuracy": 0.5716171189546585, |
| "eval_physics_num_tokens": 5021784.0, |
| "eval_physics_runtime": 52.0616, |
| "eval_physics_samples_per_second": 9.604, |
| "eval_physics_steps_per_second": 4.802, |
| "step": 300 |
| }, |
| { |
| "entropy": 1.9608435668051243, |
| "epoch": 0.248, |
| "grad_norm": 17.125, |
| "learning_rate": 6.18e-06, |
| "loss": 1.9066, |
| "mean_token_accuracy": 0.5718545202165842, |
| "num_tokens": 5201507.0, |
| "step": 310 |
| }, |
| { |
| "entropy": 1.929166065901518, |
| "epoch": 0.256, |
| "grad_norm": 16.875, |
| "learning_rate": 6.380000000000001e-06, |
| "loss": 1.8819, |
| "mean_token_accuracy": 0.5770876344293356, |
| "num_tokens": 5367308.0, |
| "step": 320 |
| }, |
| { |
| "entropy": 1.9197401903569697, |
| "epoch": 0.264, |
| "grad_norm": 18.5, |
| "learning_rate": 6.5800000000000005e-06, |
| "loss": 1.8598, |
| "mean_token_accuracy": 0.5781059630215168, |
| "num_tokens": 5536442.0, |
| "step": 330 |
| }, |
| { |
| "entropy": 1.9471155002713203, |
| "epoch": 0.272, |
| "grad_norm": 21.5, |
| "learning_rate": 6.780000000000001e-06, |
| "loss": 1.8791, |
| "mean_token_accuracy": 0.5742578504607081, |
| "num_tokens": 5699101.0, |
| "step": 340 |
| }, |
| { |
| "entropy": 1.8642883323132993, |
| "epoch": 0.28, |
| "grad_norm": 17.875, |
| "learning_rate": 6.98e-06, |
| "loss": 1.8298, |
| "mean_token_accuracy": 0.587579975090921, |
| "num_tokens": 5860398.0, |
| "step": 350 |
| }, |
| { |
| "entropy": 1.883735204488039, |
| "epoch": 0.288, |
| "grad_norm": 19.75, |
| "learning_rate": 7.180000000000001e-06, |
| "loss": 1.8306, |
| "mean_token_accuracy": 0.5885121200233698, |
| "num_tokens": 6022837.0, |
| "step": 360 |
| }, |
| { |
| "entropy": 1.832478718459606, |
| "epoch": 0.296, |
| "grad_norm": 17.0, |
| "learning_rate": 7.3800000000000005e-06, |
| "loss": 1.7789, |
| "mean_token_accuracy": 0.5987123249098658, |
| "num_tokens": 6183758.0, |
| "step": 370 |
| }, |
| { |
| "entropy": 1.8392824359238147, |
| "epoch": 0.304, |
| "grad_norm": 15.1875, |
| "learning_rate": 7.58e-06, |
| "loss": 1.7763, |
| "mean_token_accuracy": 0.595327608473599, |
| "num_tokens": 6343826.0, |
| "step": 380 |
| }, |
| { |
| "entropy": 1.8003668662160635, |
| "epoch": 0.312, |
| "grad_norm": 19.5, |
| "learning_rate": 7.78e-06, |
| "loss": 1.7664, |
| "mean_token_accuracy": 0.5983955936506391, |
| "num_tokens": 6502713.0, |
| "step": 390 |
| }, |
| { |
| "entropy": 1.7905294828116893, |
| "epoch": 0.32, |
| "grad_norm": 15.3125, |
| "learning_rate": 7.980000000000002e-06, |
| "loss": 1.7435, |
| "mean_token_accuracy": 0.6020280076190829, |
| "num_tokens": 6670164.0, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_coding_entropy": 1.8930716361999511, |
| "eval_coding_loss": 2.0930795669555664, |
| "eval_coding_mean_token_accuracy": 0.5246047481298447, |
| "eval_coding_num_tokens": 6670164.0, |
| "eval_coding_runtime": 79.1697, |
| "eval_coding_samples_per_second": 6.316, |
| "eval_coding_steps_per_second": 3.158, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.32, |
| "eval_physics_entropy": 1.7485348978042603, |
| "eval_physics_loss": 1.7591620683670044, |
| "eval_physics_mean_token_accuracy": 0.6077653738260269, |
| "eval_physics_num_tokens": 6670164.0, |
| "eval_physics_runtime": 52.1144, |
| "eval_physics_samples_per_second": 9.594, |
| "eval_physics_steps_per_second": 4.797, |
| "step": 400 |
| }, |
| { |
| "entropy": 1.791950485110283, |
| "epoch": 0.328, |
| "grad_norm": 16.25, |
| "learning_rate": 8.18e-06, |
| "loss": 1.7309, |
| "mean_token_accuracy": 0.60343619287014, |
| "num_tokens": 6831503.0, |
| "step": 410 |
| }, |
| { |
| "entropy": 1.6932228744029998, |
| "epoch": 0.336, |
| "grad_norm": 17.75, |
| "learning_rate": 8.380000000000001e-06, |
| "loss": 1.6384, |
| "mean_token_accuracy": 0.6203264124691487, |
| "num_tokens": 7003404.0, |
| "step": 420 |
| }, |
| { |
| "entropy": 1.6966097623109817, |
| "epoch": 0.344, |
| "grad_norm": 19.75, |
| "learning_rate": 8.580000000000001e-06, |
| "loss": 1.6428, |
| "mean_token_accuracy": 0.6151626098901033, |
| "num_tokens": 7172909.0, |
| "step": 430 |
| }, |
| { |
| "entropy": 1.688229451328516, |
| "epoch": 0.352, |
| "grad_norm": 18.625, |
| "learning_rate": 8.78e-06, |
| "loss": 1.6446, |
| "mean_token_accuracy": 0.619301825016737, |
| "num_tokens": 7342593.0, |
| "step": 440 |
| }, |
| { |
| "entropy": 1.684376384317875, |
| "epoch": 0.36, |
| "grad_norm": 15.1875, |
| "learning_rate": 8.98e-06, |
| "loss": 1.6384, |
| "mean_token_accuracy": 0.6227533668279648, |
| "num_tokens": 7506313.0, |
| "step": 450 |
| }, |
| { |
| "entropy": 1.6351792596280574, |
| "epoch": 0.368, |
| "grad_norm": 17.5, |
| "learning_rate": 9.180000000000002e-06, |
| "loss": 1.5883, |
| "mean_token_accuracy": 0.6302843233570457, |
| "num_tokens": 7675569.0, |
| "step": 460 |
| }, |
| { |
| "entropy": 1.6966083988547325, |
| "epoch": 0.376, |
| "grad_norm": 18.625, |
| "learning_rate": 9.38e-06, |
| "loss": 1.6569, |
| "mean_token_accuracy": 0.6154216848313808, |
| "num_tokens": 7842654.0, |
| "step": 470 |
| }, |
| { |
| "entropy": 1.6671229541301726, |
| "epoch": 0.384, |
| "grad_norm": 16.625, |
| "learning_rate": 9.58e-06, |
| "loss": 1.6255, |
| "mean_token_accuracy": 0.6238152993842959, |
| "num_tokens": 8012218.0, |
| "step": 480 |
| }, |
| { |
| "entropy": 1.634845321625471, |
| "epoch": 0.392, |
| "grad_norm": 15.0, |
| "learning_rate": 9.780000000000001e-06, |
| "loss": 1.6026, |
| "mean_token_accuracy": 0.6318565379828215, |
| "num_tokens": 8181305.0, |
| "step": 490 |
| }, |
| { |
| "entropy": 1.5764483150094748, |
| "epoch": 0.4, |
| "grad_norm": 16.125, |
| "learning_rate": 9.980000000000001e-06, |
| "loss": 1.5239, |
| "mean_token_accuracy": 0.6454750349745154, |
| "num_tokens": 8350120.0, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_coding_entropy": 1.9365736389160155, |
| "eval_coding_loss": 2.0959813594818115, |
| "eval_coding_mean_token_accuracy": 0.5250564186573029, |
| "eval_coding_num_tokens": 8350120.0, |
| "eval_coding_runtime": 79.1814, |
| "eval_coding_samples_per_second": 6.315, |
| "eval_coding_steps_per_second": 3.157, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.4, |
| "eval_physics_entropy": 1.671481309890747, |
| "eval_physics_loss": 1.6045700311660767, |
| "eval_physics_mean_token_accuracy": 0.6350576342344284, |
| "eval_physics_num_tokens": 8350120.0, |
| "eval_physics_runtime": 51.8423, |
| "eval_physics_samples_per_second": 9.645, |
| "eval_physics_steps_per_second": 4.822, |
| "step": 500 |
| }, |
| { |
| "entropy": 1.5771338425576686, |
| "epoch": 0.408, |
| "grad_norm": 15.3125, |
| "learning_rate": 1.018e-05, |
| "loss": 1.5352, |
| "mean_token_accuracy": 0.6389181109145283, |
| "num_tokens": 8520118.0, |
| "step": 510 |
| }, |
| { |
| "entropy": 1.5821339197456836, |
| "epoch": 0.416, |
| "grad_norm": 16.125, |
| "learning_rate": 1.038e-05, |
| "loss": 1.5332, |
| "mean_token_accuracy": 0.6355980787426233, |
| "num_tokens": 8688773.0, |
| "step": 520 |
| }, |
| { |
| "entropy": 1.5948798354715108, |
| "epoch": 0.424, |
| "grad_norm": 15.625, |
| "learning_rate": 1.0580000000000002e-05, |
| "loss": 1.5238, |
| "mean_token_accuracy": 0.6424462385475636, |
| "num_tokens": 8855184.0, |
| "step": 530 |
| }, |
| { |
| "entropy": 1.529571657255292, |
| "epoch": 0.432, |
| "grad_norm": 14.0625, |
| "learning_rate": 1.0780000000000002e-05, |
| "loss": 1.4919, |
| "mean_token_accuracy": 0.6440655821934342, |
| "num_tokens": 9022094.0, |
| "step": 540 |
| }, |
| { |
| "entropy": 1.5117546495050191, |
| "epoch": 0.44, |
| "grad_norm": 16.375, |
| "learning_rate": 1.0980000000000002e-05, |
| "loss": 1.4678, |
| "mean_token_accuracy": 0.6497685741633177, |
| "num_tokens": 9190509.0, |
| "step": 550 |
| }, |
| { |
| "entropy": 1.513111385703087, |
| "epoch": 0.448, |
| "grad_norm": 16.25, |
| "learning_rate": 1.1180000000000001e-05, |
| "loss": 1.4693, |
| "mean_token_accuracy": 0.6491031926125288, |
| "num_tokens": 9360608.0, |
| "step": 560 |
| }, |
| { |
| "entropy": 1.470874259248376, |
| "epoch": 0.456, |
| "grad_norm": 14.3125, |
| "learning_rate": 1.138e-05, |
| "loss": 1.415, |
| "mean_token_accuracy": 0.6601609066128731, |
| "num_tokens": 9522789.0, |
| "step": 570 |
| }, |
| { |
| "entropy": 1.5079048488289117, |
| "epoch": 0.464, |
| "grad_norm": 15.875, |
| "learning_rate": 1.1580000000000001e-05, |
| "loss": 1.4715, |
| "mean_token_accuracy": 0.6530725235119462, |
| "num_tokens": 9688896.0, |
| "step": 580 |
| }, |
| { |
| "entropy": 1.4853499613702297, |
| "epoch": 0.472, |
| "grad_norm": 14.125, |
| "learning_rate": 1.178e-05, |
| "loss": 1.4262, |
| "mean_token_accuracy": 0.6561389841139317, |
| "num_tokens": 9858464.0, |
| "step": 590 |
| }, |
| { |
| "entropy": 1.5173018813133239, |
| "epoch": 0.48, |
| "grad_norm": 21.375, |
| "learning_rate": 1.198e-05, |
| "loss": 1.4895, |
| "mean_token_accuracy": 0.6496954914182425, |
| "num_tokens": 10025864.0, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_coding_entropy": 1.953908012866974, |
| "eval_coding_loss": 2.1033880710601807, |
| "eval_coding_mean_token_accuracy": 0.5249351325035095, |
| "eval_coding_num_tokens": 10025864.0, |
| "eval_coding_runtime": 79.131, |
| "eval_coding_samples_per_second": 6.319, |
| "eval_coding_steps_per_second": 3.159, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.48, |
| "eval_physics_entropy": 1.541918823480606, |
| "eval_physics_loss": 1.4983493089675903, |
| "eval_physics_mean_token_accuracy": 0.6534626665115356, |
| "eval_physics_num_tokens": 10025864.0, |
| "eval_physics_runtime": 52.0434, |
| "eval_physics_samples_per_second": 9.607, |
| "eval_physics_steps_per_second": 4.804, |
| "step": 600 |
| }, |
| { |
| "entropy": 1.4748347599059344, |
| "epoch": 0.488, |
| "grad_norm": 16.625, |
| "learning_rate": 1.218e-05, |
| "loss": 1.4246, |
| "mean_token_accuracy": 0.6571080841124057, |
| "num_tokens": 10191232.0, |
| "step": 610 |
| }, |
| { |
| "entropy": 1.5049679655581714, |
| "epoch": 0.496, |
| "grad_norm": 14.6875, |
| "learning_rate": 1.2380000000000002e-05, |
| "loss": 1.4596, |
| "mean_token_accuracy": 0.6528839386999608, |
| "num_tokens": 10362573.0, |
| "step": 620 |
| }, |
| { |
| "entropy": 1.4375012002885341, |
| "epoch": 0.504, |
| "grad_norm": 14.4375, |
| "learning_rate": 1.2580000000000002e-05, |
| "loss": 1.3841, |
| "mean_token_accuracy": 0.6608929991722107, |
| "num_tokens": 10528450.0, |
| "step": 630 |
| }, |
| { |
| "entropy": 1.5176888063549996, |
| "epoch": 0.512, |
| "grad_norm": 14.25, |
| "learning_rate": 1.2780000000000001e-05, |
| "loss": 1.4569, |
| "mean_token_accuracy": 0.6528515750542283, |
| "num_tokens": 10694440.0, |
| "step": 640 |
| }, |
| { |
| "entropy": 1.4692929420620202, |
| "epoch": 0.52, |
| "grad_norm": 10.6875, |
| "learning_rate": 1.2980000000000001e-05, |
| "loss": 1.4449, |
| "mean_token_accuracy": 0.655676519870758, |
| "num_tokens": 10857597.0, |
| "step": 650 |
| }, |
| { |
| "entropy": 1.4483005907386541, |
| "epoch": 0.528, |
| "grad_norm": 12.125, |
| "learning_rate": 1.3180000000000001e-05, |
| "loss": 1.4046, |
| "mean_token_accuracy": 0.665894140675664, |
| "num_tokens": 11022357.0, |
| "step": 660 |
| }, |
| { |
| "entropy": 1.3869127415120601, |
| "epoch": 0.536, |
| "grad_norm": 12.8125, |
| "learning_rate": 1.3380000000000002e-05, |
| "loss": 1.3467, |
| "mean_token_accuracy": 0.672945449501276, |
| "num_tokens": 11185045.0, |
| "step": 670 |
| }, |
| { |
| "entropy": 1.4482205394655465, |
| "epoch": 0.544, |
| "grad_norm": 12.8125, |
| "learning_rate": 1.3580000000000002e-05, |
| "loss": 1.3907, |
| "mean_token_accuracy": 0.6631886938586831, |
| "num_tokens": 11345911.0, |
| "step": 680 |
| }, |
| { |
| "entropy": 1.4304094772785902, |
| "epoch": 0.552, |
| "grad_norm": 13.125, |
| "learning_rate": 1.378e-05, |
| "loss": 1.3881, |
| "mean_token_accuracy": 0.6646447636187076, |
| "num_tokens": 11518974.0, |
| "step": 690 |
| }, |
| { |
| "entropy": 1.3900580175220967, |
| "epoch": 0.56, |
| "grad_norm": 10.875, |
| "learning_rate": 1.398e-05, |
| "loss": 1.34, |
| "mean_token_accuracy": 0.6744402695447207, |
| "num_tokens": 11687225.0, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_coding_entropy": 1.8969693536758423, |
| "eval_coding_loss": 2.1377148628234863, |
| "eval_coding_mean_token_accuracy": 0.5214611231088638, |
| "eval_coding_num_tokens": 11687225.0, |
| "eval_coding_runtime": 79.1442, |
| "eval_coding_samples_per_second": 6.318, |
| "eval_coding_steps_per_second": 3.159, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.56, |
| "eval_physics_entropy": 1.3678591132164002, |
| "eval_physics_loss": 1.3962782621383667, |
| "eval_physics_mean_token_accuracy": 0.6698634608983993, |
| "eval_physics_num_tokens": 11687225.0, |
| "eval_physics_runtime": 52.0406, |
| "eval_physics_samples_per_second": 9.608, |
| "eval_physics_steps_per_second": 4.804, |
| "step": 700 |
| }, |
| { |
| "entropy": 1.3716834906488657, |
| "epoch": 0.568, |
| "grad_norm": 18.125, |
| "learning_rate": 1.418e-05, |
| "loss": 1.342, |
| "mean_token_accuracy": 0.67361662350595, |
| "num_tokens": 11848324.0, |
| "step": 710 |
| }, |
| { |
| "entropy": 1.3629539489746094, |
| "epoch": 0.576, |
| "grad_norm": 12.9375, |
| "learning_rate": 1.4380000000000001e-05, |
| "loss": 1.3293, |
| "mean_token_accuracy": 0.6735996751114726, |
| "num_tokens": 12015720.0, |
| "step": 720 |
| }, |
| { |
| "entropy": 1.3565699491649865, |
| "epoch": 0.584, |
| "grad_norm": 13.0625, |
| "learning_rate": 1.4580000000000001e-05, |
| "loss": 1.3375, |
| "mean_token_accuracy": 0.6773057371377945, |
| "num_tokens": 12178798.0, |
| "step": 730 |
| }, |
| { |
| "entropy": 1.3613810572773217, |
| "epoch": 0.592, |
| "grad_norm": 12.125, |
| "learning_rate": 1.478e-05, |
| "loss": 1.319, |
| "mean_token_accuracy": 0.6774894919246435, |
| "num_tokens": 12343393.0, |
| "step": 740 |
| }, |
| { |
| "entropy": 1.3767595831304789, |
| "epoch": 0.6, |
| "grad_norm": 11.6875, |
| "learning_rate": 1.498e-05, |
| "loss": 1.3265, |
| "mean_token_accuracy": 0.6750151777639986, |
| "num_tokens": 12509336.0, |
| "step": 750 |
| }, |
| { |
| "entropy": 1.3187911182641983, |
| "epoch": 0.608, |
| "grad_norm": 12.5, |
| "learning_rate": 1.5180000000000002e-05, |
| "loss": 1.2905, |
| "mean_token_accuracy": 0.6845489364117384, |
| "num_tokens": 12671474.0, |
| "step": 760 |
| }, |
| { |
| "entropy": 1.3427253626286983, |
| "epoch": 0.616, |
| "grad_norm": 10.75, |
| "learning_rate": 1.5380000000000002e-05, |
| "loss": 1.2963, |
| "mean_token_accuracy": 0.6806909158825875, |
| "num_tokens": 12835956.0, |
| "step": 770 |
| }, |
| { |
| "entropy": 1.3635981816798448, |
| "epoch": 0.624, |
| "grad_norm": 11.0, |
| "learning_rate": 1.5580000000000003e-05, |
| "loss": 1.3102, |
| "mean_token_accuracy": 0.6765737216919661, |
| "num_tokens": 12996077.0, |
| "step": 780 |
| }, |
| { |
| "entropy": 1.3231649551540614, |
| "epoch": 0.632, |
| "grad_norm": 12.375, |
| "learning_rate": 1.578e-05, |
| "loss": 1.2993, |
| "mean_token_accuracy": 0.6843808237463236, |
| "num_tokens": 13162966.0, |
| "step": 790 |
| }, |
| { |
| "entropy": 1.3815045531839132, |
| "epoch": 0.64, |
| "grad_norm": 11.0, |
| "learning_rate": 1.5980000000000003e-05, |
| "loss": 1.3403, |
| "mean_token_accuracy": 0.676003416068852, |
| "num_tokens": 13325291.0, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_coding_entropy": 1.8999772543907165, |
| "eval_coding_loss": 2.137219190597534, |
| "eval_coding_mean_token_accuracy": 0.5216508195400238, |
| "eval_coding_num_tokens": 13325291.0, |
| "eval_coding_runtime": 79.1227, |
| "eval_coding_samples_per_second": 6.319, |
| "eval_coding_steps_per_second": 3.16, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.64, |
| "eval_physics_entropy": 1.3213207037448884, |
| "eval_physics_loss": 1.3272991180419922, |
| "eval_physics_mean_token_accuracy": 0.6834507223367691, |
| "eval_physics_num_tokens": 13325291.0, |
| "eval_physics_runtime": 52.0498, |
| "eval_physics_samples_per_second": 9.606, |
| "eval_physics_steps_per_second": 4.803, |
| "step": 800 |
| }, |
| { |
| "entropy": 1.3278860542923212, |
| "epoch": 0.648, |
| "grad_norm": 12.0, |
| "learning_rate": 1.618e-05, |
| "loss": 1.2916, |
| "mean_token_accuracy": 0.6818249279633164, |
| "num_tokens": 13495304.0, |
| "step": 810 |
| }, |
| { |
| "entropy": 1.2705524377524853, |
| "epoch": 0.656, |
| "grad_norm": 10.875, |
| "learning_rate": 1.638e-05, |
| "loss": 1.2362, |
| "mean_token_accuracy": 0.6954903973266482, |
| "num_tokens": 13671434.0, |
| "step": 820 |
| }, |
| { |
| "entropy": 1.3271659668534994, |
| "epoch": 0.664, |
| "grad_norm": 13.375, |
| "learning_rate": 1.658e-05, |
| "loss": 1.2753, |
| "mean_token_accuracy": 0.6867586117237806, |
| "num_tokens": 13829692.0, |
| "step": 830 |
| }, |
| { |
| "entropy": 1.2718054752796888, |
| "epoch": 0.672, |
| "grad_norm": 11.125, |
| "learning_rate": 1.6780000000000002e-05, |
| "loss": 1.2239, |
| "mean_token_accuracy": 0.6929404513910413, |
| "num_tokens": 13993668.0, |
| "step": 840 |
| }, |
| { |
| "entropy": 1.2376167189329863, |
| "epoch": 0.68, |
| "grad_norm": 9.75, |
| "learning_rate": 1.698e-05, |
| "loss": 1.2134, |
| "mean_token_accuracy": 0.6992333011701704, |
| "num_tokens": 14173446.0, |
| "step": 850 |
| }, |
| { |
| "entropy": 1.3201266448944806, |
| "epoch": 0.688, |
| "grad_norm": 9.0625, |
| "learning_rate": 1.718e-05, |
| "loss": 1.2773, |
| "mean_token_accuracy": 0.6858949340879917, |
| "num_tokens": 14334907.0, |
| "step": 860 |
| }, |
| { |
| "entropy": 1.2640008773654698, |
| "epoch": 0.696, |
| "grad_norm": 8.875, |
| "learning_rate": 1.7380000000000003e-05, |
| "loss": 1.237, |
| "mean_token_accuracy": 0.6919090248644352, |
| "num_tokens": 14512555.0, |
| "step": 870 |
| }, |
| { |
| "entropy": 1.2943935077637434, |
| "epoch": 0.704, |
| "grad_norm": 11.9375, |
| "learning_rate": 1.758e-05, |
| "loss": 1.2627, |
| "mean_token_accuracy": 0.6883153811097145, |
| "num_tokens": 14670639.0, |
| "step": 880 |
| }, |
| { |
| "entropy": 1.2209337592124938, |
| "epoch": 0.712, |
| "grad_norm": 10.5, |
| "learning_rate": 1.7780000000000003e-05, |
| "loss": 1.2024, |
| "mean_token_accuracy": 0.7000889934599399, |
| "num_tokens": 14832916.0, |
| "step": 890 |
| }, |
| { |
| "entropy": 1.2753230314701796, |
| "epoch": 0.72, |
| "grad_norm": 9.0, |
| "learning_rate": 1.798e-05, |
| "loss": 1.2268, |
| "mean_token_accuracy": 0.6935238633304834, |
| "num_tokens": 14987995.0, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.72, |
| "eval_coding_entropy": 1.9145963129997254, |
| "eval_coding_loss": 2.1864311695098877, |
| "eval_coding_mean_token_accuracy": 0.5160261226892471, |
| "eval_coding_num_tokens": 14987995.0, |
| "eval_coding_runtime": 79.3569, |
| "eval_coding_samples_per_second": 6.301, |
| "eval_coding_steps_per_second": 3.15, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.72, |
| "eval_physics_entropy": 1.288757406234741, |
| "eval_physics_loss": 1.2604119777679443, |
| "eval_physics_mean_token_accuracy": 0.6937535845041275, |
| "eval_physics_num_tokens": 14987995.0, |
| "eval_physics_runtime": 52.2126, |
| "eval_physics_samples_per_second": 9.576, |
| "eval_physics_steps_per_second": 4.788, |
| "step": 900 |
| }, |
| { |
| "entropy": 1.208376468345523, |
| "epoch": 0.728, |
| "grad_norm": 10.3125, |
| "learning_rate": 1.8180000000000002e-05, |
| "loss": 1.1886, |
| "mean_token_accuracy": 0.7017698954790831, |
| "num_tokens": 15158982.0, |
| "step": 910 |
| }, |
| { |
| "entropy": 1.2472068972885608, |
| "epoch": 0.736, |
| "grad_norm": 9.6875, |
| "learning_rate": 1.8380000000000004e-05, |
| "loss": 1.2129, |
| "mean_token_accuracy": 0.6974750630557537, |
| "num_tokens": 15324659.0, |
| "step": 920 |
| }, |
| { |
| "entropy": 1.225362879037857, |
| "epoch": 0.744, |
| "grad_norm": 10.4375, |
| "learning_rate": 1.858e-05, |
| "loss": 1.183, |
| "mean_token_accuracy": 0.7002156199887395, |
| "num_tokens": 15493621.0, |
| "step": 930 |
| }, |
| { |
| "entropy": 1.2113514818251132, |
| "epoch": 0.752, |
| "grad_norm": 11.3125, |
| "learning_rate": 1.878e-05, |
| "loss": 1.1911, |
| "mean_token_accuracy": 0.7010745037347078, |
| "num_tokens": 15657342.0, |
| "step": 940 |
| }, |
| { |
| "entropy": 1.251608308404684, |
| "epoch": 0.76, |
| "grad_norm": 9.0, |
| "learning_rate": 1.898e-05, |
| "loss": 1.2108, |
| "mean_token_accuracy": 0.6982708286494017, |
| "num_tokens": 15815189.0, |
| "step": 950 |
| }, |
| { |
| "entropy": 1.2250138234347105, |
| "epoch": 0.768, |
| "grad_norm": 9.375, |
| "learning_rate": 1.918e-05, |
| "loss": 1.2033, |
| "mean_token_accuracy": 0.6997085321694613, |
| "num_tokens": 15988993.0, |
| "step": 960 |
| }, |
| { |
| "entropy": 1.2000590208917856, |
| "epoch": 0.776, |
| "grad_norm": 10.375, |
| "learning_rate": 1.938e-05, |
| "loss": 1.1522, |
| "mean_token_accuracy": 0.7049363922327757, |
| "num_tokens": 16163512.0, |
| "step": 970 |
| }, |
| { |
| "entropy": 1.2060553971678019, |
| "epoch": 0.784, |
| "grad_norm": 7.875, |
| "learning_rate": 1.9580000000000002e-05, |
| "loss": 1.1664, |
| "mean_token_accuracy": 0.7088450845330954, |
| "num_tokens": 16333857.0, |
| "step": 980 |
| }, |
| { |
| "entropy": 1.1793909084051848, |
| "epoch": 0.792, |
| "grad_norm": 7.96875, |
| "learning_rate": 1.978e-05, |
| "loss": 1.1466, |
| "mean_token_accuracy": 0.709895571321249, |
| "num_tokens": 16494118.0, |
| "step": 990 |
| }, |
| { |
| "entropy": 1.2104855172336102, |
| "epoch": 0.8, |
| "grad_norm": 10.0625, |
| "learning_rate": 1.9980000000000002e-05, |
| "loss": 1.1781, |
| "mean_token_accuracy": 0.7040123168379069, |
| "num_tokens": 16664690.0, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_coding_entropy": 1.9454177751541137, |
| "eval_coding_loss": 2.2875986099243164, |
| "eval_coding_mean_token_accuracy": 0.49295639526844026, |
| "eval_coding_num_tokens": 16664690.0, |
| "eval_coding_runtime": 79.1544, |
| "eval_coding_samples_per_second": 6.317, |
| "eval_coding_steps_per_second": 3.158, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.8, |
| "eval_physics_entropy": 1.249910979270935, |
| "eval_physics_loss": 1.215478539466858, |
| "eval_physics_mean_token_accuracy": 0.7021359713077545, |
| "eval_physics_num_tokens": 16664690.0, |
| "eval_physics_runtime": 51.8213, |
| "eval_physics_samples_per_second": 9.649, |
| "eval_physics_steps_per_second": 4.824, |
| "step": 1000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 10000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 8, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.0440286382406021e+18, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
