diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,7034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 33.222591362126245, + "eval_steps": 500, + "global_step": 10000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.03322259136212625, + "grad_norm": 16.49517250061035, + "learning_rate": 1.8e-06, + "loss": 1.1556, + "step": 10 + }, + { + "epoch": 0.0664451827242525, + "grad_norm": 5.904155731201172, + "learning_rate": 3.8e-06, + "loss": 0.9225, + "step": 20 + }, + { + "epoch": 0.09966777408637874, + "grad_norm": 3.602107048034668, + "learning_rate": 5.8e-06, + "loss": 0.4007, + "step": 30 + }, + { + "epoch": 0.132890365448505, + "grad_norm": 1.9652377367019653, + "learning_rate": 7.8e-06, + "loss": 0.2561, + "step": 40 + }, + { + "epoch": 0.16611295681063123, + "grad_norm": 2.4883804321289062, + "learning_rate": 9.800000000000001e-06, + "loss": 0.2118, + "step": 50 + }, + { + "epoch": 0.19933554817275748, + "grad_norm": 1.733266830444336, + "learning_rate": 1.18e-05, + "loss": 0.1852, + "step": 60 + }, + { + "epoch": 0.23255813953488372, + "grad_norm": 1.3866091966629028, + "learning_rate": 1.3800000000000002e-05, + "loss": 0.1448, + "step": 70 + }, + { + "epoch": 0.26578073089701, + "grad_norm": 1.1435869932174683, + "learning_rate": 1.58e-05, + "loss": 0.1502, + "step": 80 + }, + { + "epoch": 0.29900332225913623, + "grad_norm": 1.0303682088851929, + "learning_rate": 1.78e-05, + "loss": 0.1247, + "step": 90 + }, + { + "epoch": 0.33222591362126247, + "grad_norm": 1.3942605257034302, + "learning_rate": 1.9800000000000004e-05, + "loss": 0.1097, + "step": 100 + }, + { + "epoch": 0.3654485049833887, + "grad_norm": 0.9301076531410217, + "learning_rate": 2.18e-05, + "loss": 0.1061, + "step": 110 + }, + { + "epoch": 0.39867109634551495, + "grad_norm": 1.4038463830947876, + "learning_rate": 2.38e-05, + "loss": 0.09, + "step": 120 + }, + { + "epoch": 0.4318936877076412, + "grad_norm": 0.9942566752433777, + "learning_rate": 2.58e-05, + "loss": 0.0845, + "step": 130 + }, + { + "epoch": 0.46511627906976744, + "grad_norm": 0.8723886609077454, + "learning_rate": 2.7800000000000005e-05, + "loss": 0.0823, + "step": 140 + }, + { + "epoch": 0.4983388704318937, + "grad_norm": 0.7243990302085876, + "learning_rate": 2.98e-05, + "loss": 0.0692, + "step": 150 + }, + { + "epoch": 0.53156146179402, + "grad_norm": 1.035720944404602, + "learning_rate": 3.18e-05, + "loss": 0.0721, + "step": 160 + }, + { + "epoch": 0.5647840531561462, + "grad_norm": 1.8913172483444214, + "learning_rate": 3.38e-05, + "loss": 0.0761, + "step": 170 + }, + { + "epoch": 0.5980066445182725, + "grad_norm": 1.1191561222076416, + "learning_rate": 3.58e-05, + "loss": 0.081, + "step": 180 + }, + { + "epoch": 0.6312292358803987, + "grad_norm": 1.3315305709838867, + "learning_rate": 3.7800000000000004e-05, + "loss": 0.0708, + "step": 190 + }, + { + "epoch": 0.6644518272425249, + "grad_norm": 1.152005910873413, + "learning_rate": 3.9800000000000005e-05, + "loss": 0.0671, + "step": 200 + }, + { + "epoch": 0.6976744186046512, + "grad_norm": 0.9747628569602966, + "learning_rate": 4.18e-05, + "loss": 0.0678, + "step": 210 + }, + { + "epoch": 0.7308970099667774, + "grad_norm": 0.9837548732757568, + "learning_rate": 4.38e-05, + "loss": 0.0648, + "step": 220 + }, + { + "epoch": 0.7641196013289037, + "grad_norm": 1.23843514919281, + "learning_rate": 4.58e-05, + "loss": 0.0598, + "step": 230 + }, + { + "epoch": 0.7973421926910299, + "grad_norm": 1.3136545419692993, + "learning_rate": 4.78e-05, + "loss": 0.0646, + "step": 240 + }, + { + "epoch": 0.8305647840531561, + "grad_norm": 0.700904130935669, + "learning_rate": 4.9800000000000004e-05, + "loss": 0.0597, + "step": 250 + }, + { + "epoch": 0.8637873754152824, + "grad_norm": 0.6108095645904541, + "learning_rate": 5.1800000000000005e-05, + "loss": 0.0631, + "step": 260 + }, + { + "epoch": 0.8970099667774086, + "grad_norm": 0.9979882836341858, + "learning_rate": 5.380000000000001e-05, + "loss": 0.0654, + "step": 270 + }, + { + "epoch": 0.9302325581395349, + "grad_norm": 0.8695108890533447, + "learning_rate": 5.580000000000001e-05, + "loss": 0.0621, + "step": 280 + }, + { + "epoch": 0.9634551495016611, + "grad_norm": 0.5228680968284607, + "learning_rate": 5.7799999999999995e-05, + "loss": 0.0615, + "step": 290 + }, + { + "epoch": 0.9966777408637874, + "grad_norm": 0.7989741563796997, + "learning_rate": 5.9800000000000003e-05, + "loss": 0.0572, + "step": 300 + }, + { + "epoch": 1.0299003322259137, + "grad_norm": 1.0117868185043335, + "learning_rate": 6.18e-05, + "loss": 0.0832, + "step": 310 + }, + { + "epoch": 1.06312292358804, + "grad_norm": 0.6062241792678833, + "learning_rate": 6.38e-05, + "loss": 0.0601, + "step": 320 + }, + { + "epoch": 1.0963455149501662, + "grad_norm": 0.8119208216667175, + "learning_rate": 6.58e-05, + "loss": 0.061, + "step": 330 + }, + { + "epoch": 1.1295681063122924, + "grad_norm": 1.240936517715454, + "learning_rate": 6.780000000000001e-05, + "loss": 0.0579, + "step": 340 + }, + { + "epoch": 1.1627906976744187, + "grad_norm": 0.6764455437660217, + "learning_rate": 6.98e-05, + "loss": 0.0593, + "step": 350 + }, + { + "epoch": 1.196013289036545, + "grad_norm": 0.9562680721282959, + "learning_rate": 7.18e-05, + "loss": 0.0566, + "step": 360 + }, + { + "epoch": 1.2292358803986712, + "grad_norm": 0.909827709197998, + "learning_rate": 7.38e-05, + "loss": 0.0601, + "step": 370 + }, + { + "epoch": 1.2624584717607974, + "grad_norm": 0.9948165416717529, + "learning_rate": 7.58e-05, + "loss": 0.0564, + "step": 380 + }, + { + "epoch": 1.2956810631229236, + "grad_norm": 0.8981932997703552, + "learning_rate": 7.780000000000001e-05, + "loss": 0.0579, + "step": 390 + }, + { + "epoch": 1.3289036544850499, + "grad_norm": 1.1012845039367676, + "learning_rate": 7.98e-05, + "loss": 0.0563, + "step": 400 + }, + { + "epoch": 1.3621262458471761, + "grad_norm": 0.6229888200759888, + "learning_rate": 8.18e-05, + "loss": 0.0544, + "step": 410 + }, + { + "epoch": 1.3953488372093024, + "grad_norm": 0.6055900454521179, + "learning_rate": 8.38e-05, + "loss": 0.0491, + "step": 420 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.662727952003479, + "learning_rate": 8.58e-05, + "loss": 0.0526, + "step": 430 + }, + { + "epoch": 1.4617940199335548, + "grad_norm": 0.5412904024124146, + "learning_rate": 8.78e-05, + "loss": 0.0524, + "step": 440 + }, + { + "epoch": 1.495016611295681, + "grad_norm": 0.552821934223175, + "learning_rate": 8.98e-05, + "loss": 0.0518, + "step": 450 + }, + { + "epoch": 1.5282392026578073, + "grad_norm": 0.7804833650588989, + "learning_rate": 9.180000000000001e-05, + "loss": 0.0509, + "step": 460 + }, + { + "epoch": 1.5614617940199336, + "grad_norm": 0.6241787672042847, + "learning_rate": 9.38e-05, + "loss": 0.0506, + "step": 470 + }, + { + "epoch": 1.5946843853820598, + "grad_norm": 0.6631738543510437, + "learning_rate": 9.58e-05, + "loss": 0.0489, + "step": 480 + }, + { + "epoch": 1.627906976744186, + "grad_norm": 0.5631969571113586, + "learning_rate": 9.78e-05, + "loss": 0.0399, + "step": 490 + }, + { + "epoch": 1.6611295681063123, + "grad_norm": 0.6725241541862488, + "learning_rate": 9.98e-05, + "loss": 0.0493, + "step": 500 + }, + { + "epoch": 1.6943521594684385, + "grad_norm": 0.7589104175567627, + "learning_rate": 9.9999778549206e-05, + "loss": 0.047, + "step": 510 + }, + { + "epoch": 1.7275747508305648, + "grad_norm": 0.7991833090782166, + "learning_rate": 9.999901304280685e-05, + "loss": 0.0473, + "step": 520 + }, + { + "epoch": 1.760797342192691, + "grad_norm": 0.7787967920303345, + "learning_rate": 9.999770075521164e-05, + "loss": 0.0494, + "step": 530 + }, + { + "epoch": 1.7940199335548173, + "grad_norm": 0.6019904613494873, + "learning_rate": 9.99958417007713e-05, + "loss": 0.048, + "step": 540 + }, + { + "epoch": 1.8272425249169435, + "grad_norm": 0.5640949606895447, + "learning_rate": 9.999343589981615e-05, + "loss": 0.0454, + "step": 550 + }, + { + "epoch": 1.8604651162790697, + "grad_norm": 0.44856587052345276, + "learning_rate": 9.999048337865568e-05, + "loss": 0.0455, + "step": 560 + }, + { + "epoch": 1.893687707641196, + "grad_norm": 0.6371676921844482, + "learning_rate": 9.998698416957815e-05, + "loss": 0.0429, + "step": 570 + }, + { + "epoch": 1.9269102990033222, + "grad_norm": 0.8966260552406311, + "learning_rate": 9.998293831085037e-05, + "loss": 0.048, + "step": 580 + }, + { + "epoch": 1.9601328903654485, + "grad_norm": 0.5906923413276672, + "learning_rate": 9.997834584671719e-05, + "loss": 0.046, + "step": 590 + }, + { + "epoch": 1.9933554817275747, + "grad_norm": 0.5524625778198242, + "learning_rate": 9.997320682740107e-05, + "loss": 0.0476, + "step": 600 + }, + { + "epoch": 2.026578073089701, + "grad_norm": 0.5328683853149414, + "learning_rate": 9.996752130910149e-05, + "loss": 0.0485, + "step": 610 + }, + { + "epoch": 2.0598006644518274, + "grad_norm": 0.7532830834388733, + "learning_rate": 9.99612893539944e-05, + "loss": 0.0429, + "step": 620 + }, + { + "epoch": 2.0930232558139537, + "grad_norm": 0.5240247845649719, + "learning_rate": 9.995451103023144e-05, + "loss": 0.0367, + "step": 630 + }, + { + "epoch": 2.12624584717608, + "grad_norm": 0.5653980374336243, + "learning_rate": 9.994718641193928e-05, + "loss": 0.0396, + "step": 640 + }, + { + "epoch": 2.159468438538206, + "grad_norm": 0.7227243185043335, + "learning_rate": 9.993931557921874e-05, + "loss": 0.0396, + "step": 650 + }, + { + "epoch": 2.1926910299003324, + "grad_norm": 0.5951940417289734, + "learning_rate": 9.993089861814402e-05, + "loss": 0.0412, + "step": 660 + }, + { + "epoch": 2.2259136212624586, + "grad_norm": 0.5065795183181763, + "learning_rate": 9.992193562076166e-05, + "loss": 0.0414, + "step": 670 + }, + { + "epoch": 2.259136212624585, + "grad_norm": 0.5497207045555115, + "learning_rate": 9.991242668508954e-05, + "loss": 0.0426, + "step": 680 + }, + { + "epoch": 2.292358803986711, + "grad_norm": 0.462469220161438, + "learning_rate": 9.990237191511587e-05, + "loss": 0.0421, + "step": 690 + }, + { + "epoch": 2.3255813953488373, + "grad_norm": 0.6984947919845581, + "learning_rate": 9.989177142079802e-05, + "loss": 0.0446, + "step": 700 + }, + { + "epoch": 2.3588039867109636, + "grad_norm": 0.443020224571228, + "learning_rate": 9.988062531806126e-05, + "loss": 0.0462, + "step": 710 + }, + { + "epoch": 2.39202657807309, + "grad_norm": 0.5449220538139343, + "learning_rate": 9.986893372879762e-05, + "loss": 0.0446, + "step": 720 + }, + { + "epoch": 2.425249169435216, + "grad_norm": 0.5548967123031616, + "learning_rate": 9.985669678086443e-05, + "loss": 0.0391, + "step": 730 + }, + { + "epoch": 2.4584717607973423, + "grad_norm": 0.8857158422470093, + "learning_rate": 9.984391460808298e-05, + "loss": 0.0382, + "step": 740 + }, + { + "epoch": 2.4916943521594686, + "grad_norm": 0.8469329476356506, + "learning_rate": 9.983058735023709e-05, + "loss": 0.0362, + "step": 750 + }, + { + "epoch": 2.524916943521595, + "grad_norm": 0.4711757004261017, + "learning_rate": 9.98167151530715e-05, + "loss": 0.034, + "step": 760 + }, + { + "epoch": 2.558139534883721, + "grad_norm": 0.4842841923236847, + "learning_rate": 9.980229816829034e-05, + "loss": 0.0298, + "step": 770 + }, + { + "epoch": 2.5913621262458473, + "grad_norm": 0.8352331519126892, + "learning_rate": 9.978733655355544e-05, + "loss": 0.0368, + "step": 780 + }, + { + "epoch": 2.6245847176079735, + "grad_norm": 0.7118296027183533, + "learning_rate": 9.977183047248464e-05, + "loss": 0.033, + "step": 790 + }, + { + "epoch": 2.6578073089700998, + "grad_norm": 0.36444178223609924, + "learning_rate": 9.975578009464992e-05, + "loss": 0.0402, + "step": 800 + }, + { + "epoch": 2.691029900332226, + "grad_norm": 0.4943622052669525, + "learning_rate": 9.97391855955757e-05, + "loss": 0.0397, + "step": 810 + }, + { + "epoch": 2.7242524916943522, + "grad_norm": 0.5186375975608826, + "learning_rate": 9.972204715673669e-05, + "loss": 0.0391, + "step": 820 + }, + { + "epoch": 2.7574750830564785, + "grad_norm": 0.4055125117301941, + "learning_rate": 9.970436496555617e-05, + "loss": 0.0373, + "step": 830 + }, + { + "epoch": 2.7906976744186047, + "grad_norm": 0.5435572862625122, + "learning_rate": 9.968613921540373e-05, + "loss": 0.0321, + "step": 840 + }, + { + "epoch": 2.823920265780731, + "grad_norm": 0.5882667303085327, + "learning_rate": 9.966737010559326e-05, + "loss": 0.0372, + "step": 850 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.6232360601425171, + "learning_rate": 9.964805784138072e-05, + "loss": 0.0317, + "step": 860 + }, + { + "epoch": 2.8903654485049834, + "grad_norm": 0.47762417793273926, + "learning_rate": 9.962820263396195e-05, + "loss": 0.0345, + "step": 870 + }, + { + "epoch": 2.9235880398671097, + "grad_norm": 0.645271360874176, + "learning_rate": 9.960780470047033e-05, + "loss": 0.0338, + "step": 880 + }, + { + "epoch": 2.956810631229236, + "grad_norm": 0.6059306263923645, + "learning_rate": 9.958686426397437e-05, + "loss": 0.0326, + "step": 890 + }, + { + "epoch": 2.990033222591362, + "grad_norm": 0.4821760356426239, + "learning_rate": 9.956538155347534e-05, + "loss": 0.0374, + "step": 900 + }, + { + "epoch": 3.0232558139534884, + "grad_norm": 0.6130004525184631, + "learning_rate": 9.95433568039047e-05, + "loss": 0.0283, + "step": 910 + }, + { + "epoch": 3.0564784053156147, + "grad_norm": 0.5384393930435181, + "learning_rate": 9.952079025612162e-05, + "loss": 0.0286, + "step": 920 + }, + { + "epoch": 3.089700996677741, + "grad_norm": 0.6059784293174744, + "learning_rate": 9.949768215691022e-05, + "loss": 0.0261, + "step": 930 + }, + { + "epoch": 3.122923588039867, + "grad_norm": 0.6603447794914246, + "learning_rate": 9.9474032758977e-05, + "loss": 0.0301, + "step": 940 + }, + { + "epoch": 3.1561461794019934, + "grad_norm": 0.5607938766479492, + "learning_rate": 9.944984232094794e-05, + "loss": 0.033, + "step": 950 + }, + { + "epoch": 3.1893687707641196, + "grad_norm": 0.3325658142566681, + "learning_rate": 9.942511110736584e-05, + "loss": 0.0352, + "step": 960 + }, + { + "epoch": 3.222591362126246, + "grad_norm": 0.5348853468894958, + "learning_rate": 9.939983938868726e-05, + "loss": 0.0312, + "step": 970 + }, + { + "epoch": 3.255813953488372, + "grad_norm": 0.5491706728935242, + "learning_rate": 9.93740274412797e-05, + "loss": 0.0279, + "step": 980 + }, + { + "epoch": 3.2890365448504983, + "grad_norm": 0.3640613257884979, + "learning_rate": 9.934767554741846e-05, + "loss": 0.0292, + "step": 990 + }, + { + "epoch": 3.3222591362126246, + "grad_norm": 0.40242475271224976, + "learning_rate": 9.932078399528361e-05, + "loss": 0.0269, + "step": 1000 + }, + { + "epoch": 3.355481727574751, + "grad_norm": 0.3187004625797272, + "learning_rate": 9.929335307895689e-05, + "loss": 0.0278, + "step": 1010 + }, + { + "epoch": 3.388704318936877, + "grad_norm": 0.5363311767578125, + "learning_rate": 9.926538309841839e-05, + "loss": 0.0266, + "step": 1020 + }, + { + "epoch": 3.4219269102990033, + "grad_norm": 0.5733344554901123, + "learning_rate": 9.923687435954334e-05, + "loss": 0.0305, + "step": 1030 + }, + { + "epoch": 3.4551495016611296, + "grad_norm": 0.4922335147857666, + "learning_rate": 9.920782717409873e-05, + "loss": 0.0307, + "step": 1040 + }, + { + "epoch": 3.488372093023256, + "grad_norm": 0.6312623620033264, + "learning_rate": 9.917824185973994e-05, + "loss": 0.0281, + "step": 1050 + }, + { + "epoch": 3.521594684385382, + "grad_norm": 0.4072914719581604, + "learning_rate": 9.914811874000723e-05, + "loss": 0.0347, + "step": 1060 + }, + { + "epoch": 3.5548172757475083, + "grad_norm": 0.386290580034256, + "learning_rate": 9.911745814432218e-05, + "loss": 0.0397, + "step": 1070 + }, + { + "epoch": 3.5880398671096345, + "grad_norm": 0.4319952130317688, + "learning_rate": 9.90862604079842e-05, + "loss": 0.028, + "step": 1080 + }, + { + "epoch": 3.6212624584717608, + "grad_norm": 0.3090722858905792, + "learning_rate": 9.90545258721667e-05, + "loss": 0.0294, + "step": 1090 + }, + { + "epoch": 3.654485049833887, + "grad_norm": 0.41966256499290466, + "learning_rate": 9.90222548839135e-05, + "loss": 0.0283, + "step": 1100 + }, + { + "epoch": 3.6877076411960132, + "grad_norm": 0.34183767437934875, + "learning_rate": 9.898944779613495e-05, + "loss": 0.0311, + "step": 1110 + }, + { + "epoch": 3.7209302325581395, + "grad_norm": 0.29612553119659424, + "learning_rate": 9.89561049676041e-05, + "loss": 0.0263, + "step": 1120 + }, + { + "epoch": 3.7541528239202657, + "grad_norm": 0.29920074343681335, + "learning_rate": 9.89222267629528e-05, + "loss": 0.0315, + "step": 1130 + }, + { + "epoch": 3.787375415282392, + "grad_norm": 0.4629840552806854, + "learning_rate": 9.888781355266763e-05, + "loss": 0.0297, + "step": 1140 + }, + { + "epoch": 3.820598006644518, + "grad_norm": 0.6953020691871643, + "learning_rate": 9.885286571308598e-05, + "loss": 0.0279, + "step": 1150 + }, + { + "epoch": 3.8538205980066444, + "grad_norm": 0.4963432848453522, + "learning_rate": 9.881738362639182e-05, + "loss": 0.0276, + "step": 1160 + }, + { + "epoch": 3.8870431893687707, + "grad_norm": 0.1942342221736908, + "learning_rate": 9.878136768061154e-05, + "loss": 0.0275, + "step": 1170 + }, + { + "epoch": 3.920265780730897, + "grad_norm": 0.46550989151000977, + "learning_rate": 9.874481826960979e-05, + "loss": 0.0278, + "step": 1180 + }, + { + "epoch": 3.953488372093023, + "grad_norm": 0.5046531558036804, + "learning_rate": 9.870773579308503e-05, + "loss": 0.0292, + "step": 1190 + }, + { + "epoch": 3.9867109634551494, + "grad_norm": 0.7471413016319275, + "learning_rate": 9.867012065656533e-05, + "loss": 0.0302, + "step": 1200 + }, + { + "epoch": 4.019933554817276, + "grad_norm": 0.4875882863998413, + "learning_rate": 9.863197327140376e-05, + "loss": 0.0321, + "step": 1210 + }, + { + "epoch": 4.053156146179402, + "grad_norm": 0.4417405426502228, + "learning_rate": 9.859329405477403e-05, + "loss": 0.0303, + "step": 1220 + }, + { + "epoch": 4.086378737541528, + "grad_norm": 0.3285420835018158, + "learning_rate": 9.855408342966585e-05, + "loss": 0.0308, + "step": 1230 + }, + { + "epoch": 4.119601328903655, + "grad_norm": 0.383506178855896, + "learning_rate": 9.851434182488033e-05, + "loss": 0.0308, + "step": 1240 + }, + { + "epoch": 4.152823920265781, + "grad_norm": 0.4398779571056366, + "learning_rate": 9.84740696750253e-05, + "loss": 0.0292, + "step": 1250 + }, + { + "epoch": 4.186046511627907, + "grad_norm": 0.5198376178741455, + "learning_rate": 9.843326742051055e-05, + "loss": 0.0325, + "step": 1260 + }, + { + "epoch": 4.219269102990033, + "grad_norm": 0.4738619923591614, + "learning_rate": 9.839193550754297e-05, + "loss": 0.0251, + "step": 1270 + }, + { + "epoch": 4.25249169435216, + "grad_norm": 0.33822280168533325, + "learning_rate": 9.835007438812177e-05, + "loss": 0.0312, + "step": 1280 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.500541090965271, + "learning_rate": 9.830768452003341e-05, + "loss": 0.0348, + "step": 1290 + }, + { + "epoch": 4.318936877076412, + "grad_norm": 0.3859885632991791, + "learning_rate": 9.826476636684671e-05, + "loss": 0.0303, + "step": 1300 + }, + { + "epoch": 4.352159468438538, + "grad_norm": 0.43077021837234497, + "learning_rate": 9.822132039790773e-05, + "loss": 0.0276, + "step": 1310 + }, + { + "epoch": 4.385382059800665, + "grad_norm": 0.4338962435722351, + "learning_rate": 9.817734708833461e-05, + "loss": 0.0272, + "step": 1320 + }, + { + "epoch": 4.4186046511627906, + "grad_norm": 0.5305380821228027, + "learning_rate": 9.813284691901243e-05, + "loss": 0.0224, + "step": 1330 + }, + { + "epoch": 4.451827242524917, + "grad_norm": 0.181693434715271, + "learning_rate": 9.808782037658792e-05, + "loss": 0.024, + "step": 1340 + }, + { + "epoch": 4.485049833887043, + "grad_norm": 0.35233795642852783, + "learning_rate": 9.804226795346411e-05, + "loss": 0.0244, + "step": 1350 + }, + { + "epoch": 4.51827242524917, + "grad_norm": 0.5381154417991638, + "learning_rate": 9.799619014779503e-05, + "loss": 0.0261, + "step": 1360 + }, + { + "epoch": 4.5514950166112955, + "grad_norm": 0.43545234203338623, + "learning_rate": 9.794958746348013e-05, + "loss": 0.0238, + "step": 1370 + }, + { + "epoch": 4.584717607973422, + "grad_norm": 0.2619587481021881, + "learning_rate": 9.790246041015896e-05, + "loss": 0.0265, + "step": 1380 + }, + { + "epoch": 4.617940199335548, + "grad_norm": 0.3800243139266968, + "learning_rate": 9.785480950320538e-05, + "loss": 0.0227, + "step": 1390 + }, + { + "epoch": 4.651162790697675, + "grad_norm": 0.38735443353652954, + "learning_rate": 9.78066352637221e-05, + "loss": 0.0259, + "step": 1400 + }, + { + "epoch": 4.6843853820598005, + "grad_norm": 0.4869486689567566, + "learning_rate": 9.775793821853488e-05, + "loss": 0.0269, + "step": 1410 + }, + { + "epoch": 4.717607973421927, + "grad_norm": 0.23203584551811218, + "learning_rate": 9.77087189001868e-05, + "loss": 0.0303, + "step": 1420 + }, + { + "epoch": 4.750830564784053, + "grad_norm": 0.37429264187812805, + "learning_rate": 9.765897784693243e-05, + "loss": 0.0257, + "step": 1430 + }, + { + "epoch": 4.78405315614618, + "grad_norm": 0.5015078186988831, + "learning_rate": 9.760871560273197e-05, + "loss": 0.032, + "step": 1440 + }, + { + "epoch": 4.8172757475083055, + "grad_norm": 0.3568253815174103, + "learning_rate": 9.755793271724526e-05, + "loss": 0.0277, + "step": 1450 + }, + { + "epoch": 4.850498338870432, + "grad_norm": 0.3456842303276062, + "learning_rate": 9.750662974582584e-05, + "loss": 0.0249, + "step": 1460 + }, + { + "epoch": 4.883720930232558, + "grad_norm": 0.3531418442726135, + "learning_rate": 9.745480724951473e-05, + "loss": 0.0245, + "step": 1470 + }, + { + "epoch": 4.916943521594685, + "grad_norm": 0.2724597752094269, + "learning_rate": 9.740246579503447e-05, + "loss": 0.0305, + "step": 1480 + }, + { + "epoch": 4.95016611295681, + "grad_norm": 0.2732029855251312, + "learning_rate": 9.734960595478284e-05, + "loss": 0.0233, + "step": 1490 + }, + { + "epoch": 4.983388704318937, + "grad_norm": 0.3414079546928406, + "learning_rate": 9.729622830682657e-05, + "loss": 0.0248, + "step": 1500 + }, + { + "epoch": 5.016611295681063, + "grad_norm": 0.5630635619163513, + "learning_rate": 9.724233343489504e-05, + "loss": 0.0329, + "step": 1510 + }, + { + "epoch": 5.04983388704319, + "grad_norm": 0.4476262629032135, + "learning_rate": 9.718792192837396e-05, + "loss": 0.0254, + "step": 1520 + }, + { + "epoch": 5.083056478405315, + "grad_norm": 0.3482656478881836, + "learning_rate": 9.713299438229886e-05, + "loss": 0.03, + "step": 1530 + }, + { + "epoch": 5.116279069767442, + "grad_norm": 0.42722585797309875, + "learning_rate": 9.707755139734855e-05, + "loss": 0.0226, + "step": 1540 + }, + { + "epoch": 5.149501661129568, + "grad_norm": 0.44865331053733826, + "learning_rate": 9.702159357983866e-05, + "loss": 0.035, + "step": 1550 + }, + { + "epoch": 5.1827242524916945, + "grad_norm": 0.3667987287044525, + "learning_rate": 9.696512154171492e-05, + "loss": 0.025, + "step": 1560 + }, + { + "epoch": 5.21594684385382, + "grad_norm": 0.2647327780723572, + "learning_rate": 9.690813590054645e-05, + "loss": 0.0268, + "step": 1570 + }, + { + "epoch": 5.249169435215947, + "grad_norm": 0.32716986536979675, + "learning_rate": 9.685063727951914e-05, + "loss": 0.0242, + "step": 1580 + }, + { + "epoch": 5.282392026578073, + "grad_norm": 0.32449787855148315, + "learning_rate": 9.679262630742865e-05, + "loss": 0.0233, + "step": 1590 + }, + { + "epoch": 5.3156146179401995, + "grad_norm": 0.417737752199173, + "learning_rate": 9.673410361867373e-05, + "loss": 0.0251, + "step": 1600 + }, + { + "epoch": 5.348837209302325, + "grad_norm": 0.33530154824256897, + "learning_rate": 9.667506985324909e-05, + "loss": 0.0238, + "step": 1610 + }, + { + "epoch": 5.382059800664452, + "grad_norm": 0.39416244626045227, + "learning_rate": 9.661552565673855e-05, + "loss": 0.0251, + "step": 1620 + }, + { + "epoch": 5.415282392026578, + "grad_norm": 0.3443300724029541, + "learning_rate": 9.655547168030789e-05, + "loss": 0.0218, + "step": 1630 + }, + { + "epoch": 5.4485049833887045, + "grad_norm": 0.4587835371494293, + "learning_rate": 9.649490858069777e-05, + "loss": 0.0258, + "step": 1640 + }, + { + "epoch": 5.48172757475083, + "grad_norm": 0.41582098603248596, + "learning_rate": 9.643383702021658e-05, + "loss": 0.0261, + "step": 1650 + }, + { + "epoch": 5.514950166112957, + "grad_norm": 0.28211021423339844, + "learning_rate": 9.637225766673307e-05, + "loss": 0.0267, + "step": 1660 + }, + { + "epoch": 5.548172757475083, + "grad_norm": 0.4471738934516907, + "learning_rate": 9.631017119366922e-05, + "loss": 0.0235, + "step": 1670 + }, + { + "epoch": 5.5813953488372094, + "grad_norm": 0.28704577684402466, + "learning_rate": 9.624757827999273e-05, + "loss": 0.0233, + "step": 1680 + }, + { + "epoch": 5.614617940199335, + "grad_norm": 0.41606009006500244, + "learning_rate": 9.618447961020971e-05, + "loss": 0.0274, + "step": 1690 + }, + { + "epoch": 5.647840531561462, + "grad_norm": 0.47395747900009155, + "learning_rate": 9.612087587435707e-05, + "loss": 0.0224, + "step": 1700 + }, + { + "epoch": 5.681063122923588, + "grad_norm": 0.347327321767807, + "learning_rate": 9.605676776799508e-05, + "loss": 0.0218, + "step": 1710 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.2327083796262741, + "learning_rate": 9.599215599219973e-05, + "loss": 0.0234, + "step": 1720 + }, + { + "epoch": 5.74750830564784, + "grad_norm": 0.3055405020713806, + "learning_rate": 9.592704125355505e-05, + "loss": 0.0242, + "step": 1730 + }, + { + "epoch": 5.780730897009967, + "grad_norm": 0.46758589148521423, + "learning_rate": 9.586142426414538e-05, + "loss": 0.0236, + "step": 1740 + }, + { + "epoch": 5.813953488372093, + "grad_norm": 0.42915889620780945, + "learning_rate": 9.57953057415476e-05, + "loss": 0.0236, + "step": 1750 + }, + { + "epoch": 5.847176079734219, + "grad_norm": 0.3505151867866516, + "learning_rate": 9.572868640882328e-05, + "loss": 0.0226, + "step": 1760 + }, + { + "epoch": 5.880398671096345, + "grad_norm": 0.3558598756790161, + "learning_rate": 9.56615669945108e-05, + "loss": 0.0238, + "step": 1770 + }, + { + "epoch": 5.913621262458472, + "grad_norm": 0.32505741715431213, + "learning_rate": 9.55939482326173e-05, + "loss": 0.0274, + "step": 1780 + }, + { + "epoch": 5.946843853820598, + "grad_norm": 0.3497382402420044, + "learning_rate": 9.552583086261069e-05, + "loss": 0.0288, + "step": 1790 + }, + { + "epoch": 5.980066445182724, + "grad_norm": 0.336085706949234, + "learning_rate": 9.545721562941168e-05, + "loss": 0.0285, + "step": 1800 + }, + { + "epoch": 6.01328903654485, + "grad_norm": 0.4602932929992676, + "learning_rate": 9.538810328338543e-05, + "loss": 0.0232, + "step": 1810 + }, + { + "epoch": 6.046511627906977, + "grad_norm": 0.37610024213790894, + "learning_rate": 9.531849458033349e-05, + "loss": 0.0304, + "step": 1820 + }, + { + "epoch": 6.079734219269103, + "grad_norm": 0.4268096685409546, + "learning_rate": 9.524839028148547e-05, + "loss": 0.0243, + "step": 1830 + }, + { + "epoch": 6.112956810631229, + "grad_norm": 0.705152153968811, + "learning_rate": 9.517779115349077e-05, + "loss": 0.0231, + "step": 1840 + }, + { + "epoch": 6.146179401993355, + "grad_norm": 0.3031271696090698, + "learning_rate": 9.510669796841014e-05, + "loss": 0.0249, + "step": 1850 + }, + { + "epoch": 6.179401993355482, + "grad_norm": 0.46579620242118835, + "learning_rate": 9.503511150370727e-05, + "loss": 0.025, + "step": 1860 + }, + { + "epoch": 6.212624584717608, + "grad_norm": 0.4709901511669159, + "learning_rate": 9.496303254224024e-05, + "loss": 0.03, + "step": 1870 + }, + { + "epoch": 6.245847176079734, + "grad_norm": 0.20217536389827728, + "learning_rate": 9.489046187225306e-05, + "loss": 0.027, + "step": 1880 + }, + { + "epoch": 6.27906976744186, + "grad_norm": 0.35509705543518066, + "learning_rate": 9.481740028736692e-05, + "loss": 0.0183, + "step": 1890 + }, + { + "epoch": 6.312292358803987, + "grad_norm": 0.299204558134079, + "learning_rate": 9.474384858657164e-05, + "loss": 0.0241, + "step": 1900 + }, + { + "epoch": 6.3455149501661126, + "grad_norm": 0.3697379529476166, + "learning_rate": 9.466980757421679e-05, + "loss": 0.0269, + "step": 1910 + }, + { + "epoch": 6.378737541528239, + "grad_norm": 0.5883201956748962, + "learning_rate": 9.459527806000305e-05, + "loss": 0.0218, + "step": 1920 + }, + { + "epoch": 6.411960132890365, + "grad_norm": 0.39769867062568665, + "learning_rate": 9.452026085897325e-05, + "loss": 0.0245, + "step": 1930 + }, + { + "epoch": 6.445182724252492, + "grad_norm": 0.5797551274299622, + "learning_rate": 9.444475679150348e-05, + "loss": 0.0265, + "step": 1940 + }, + { + "epoch": 6.4784053156146175, + "grad_norm": 0.462607204914093, + "learning_rate": 9.436876668329411e-05, + "loss": 0.0255, + "step": 1950 + }, + { + "epoch": 6.511627906976744, + "grad_norm": 0.41783976554870605, + "learning_rate": 9.429229136536079e-05, + "loss": 0.0228, + "step": 1960 + }, + { + "epoch": 6.544850498338871, + "grad_norm": 0.31457412242889404, + "learning_rate": 9.421533167402534e-05, + "loss": 0.0233, + "step": 1970 + }, + { + "epoch": 6.578073089700997, + "grad_norm": 0.28764960169792175, + "learning_rate": 9.413788845090666e-05, + "loss": 0.023, + "step": 1980 + }, + { + "epoch": 6.6112956810631225, + "grad_norm": 0.2633940875530243, + "learning_rate": 9.405996254291136e-05, + "loss": 0.0264, + "step": 1990 + }, + { + "epoch": 6.644518272425249, + "grad_norm": 0.33487606048583984, + "learning_rate": 9.398155480222474e-05, + "loss": 0.023, + "step": 2000 + }, + { + "epoch": 6.677740863787376, + "grad_norm": 0.3526790142059326, + "learning_rate": 9.390266608630128e-05, + "loss": 0.0212, + "step": 2010 + }, + { + "epoch": 6.710963455149502, + "grad_norm": 0.4186701774597168, + "learning_rate": 9.38232972578553e-05, + "loss": 0.0218, + "step": 2020 + }, + { + "epoch": 6.7441860465116275, + "grad_norm": 0.36500057578086853, + "learning_rate": 9.374344918485164e-05, + "loss": 0.0267, + "step": 2030 + }, + { + "epoch": 6.777408637873754, + "grad_norm": 0.4534339904785156, + "learning_rate": 9.366312274049602e-05, + "loss": 0.0271, + "step": 2040 + }, + { + "epoch": 6.810631229235881, + "grad_norm": 0.32824012637138367, + "learning_rate": 9.358231880322554e-05, + "loss": 0.0232, + "step": 2050 + }, + { + "epoch": 6.843853820598007, + "grad_norm": 0.37586942315101624, + "learning_rate": 9.350103825669916e-05, + "loss": 0.0202, + "step": 2060 + }, + { + "epoch": 6.877076411960132, + "grad_norm": 0.21727822721004486, + "learning_rate": 9.341928198978787e-05, + "loss": 0.0197, + "step": 2070 + }, + { + "epoch": 6.910299003322259, + "grad_norm": 0.3569616377353668, + "learning_rate": 9.333705089656512e-05, + "loss": 0.0223, + "step": 2080 + }, + { + "epoch": 6.943521594684386, + "grad_norm": 0.3130171000957489, + "learning_rate": 9.325434587629698e-05, + "loss": 0.0198, + "step": 2090 + }, + { + "epoch": 6.976744186046512, + "grad_norm": 0.3402811288833618, + "learning_rate": 9.31711678334323e-05, + "loss": 0.0224, + "step": 2100 + }, + { + "epoch": 7.009966777408638, + "grad_norm": 0.3928144872188568, + "learning_rate": 9.308751767759282e-05, + "loss": 0.0184, + "step": 2110 + }, + { + "epoch": 7.043189368770764, + "grad_norm": 0.39649125933647156, + "learning_rate": 9.300339632356325e-05, + "loss": 0.0206, + "step": 2120 + }, + { + "epoch": 7.076411960132891, + "grad_norm": 0.5337740778923035, + "learning_rate": 9.291880469128124e-05, + "loss": 0.023, + "step": 2130 + }, + { + "epoch": 7.1096345514950166, + "grad_norm": 0.4840176999568939, + "learning_rate": 9.283374370582732e-05, + "loss": 0.0217, + "step": 2140 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.3124825656414032, + "learning_rate": 9.274821429741482e-05, + "loss": 0.0252, + "step": 2150 + }, + { + "epoch": 7.176079734219269, + "grad_norm": 0.38995927572250366, + "learning_rate": 9.266221740137961e-05, + "loss": 0.0193, + "step": 2160 + }, + { + "epoch": 7.209302325581396, + "grad_norm": 0.4084652066230774, + "learning_rate": 9.257575395817001e-05, + "loss": 0.0235, + "step": 2170 + }, + { + "epoch": 7.2425249169435215, + "grad_norm": 0.189564511179924, + "learning_rate": 9.248882491333637e-05, + "loss": 0.0176, + "step": 2180 + }, + { + "epoch": 7.275747508305648, + "grad_norm": 0.38772428035736084, + "learning_rate": 9.240143121752076e-05, + "loss": 0.0215, + "step": 2190 + }, + { + "epoch": 7.308970099667774, + "grad_norm": 0.4078693985939026, + "learning_rate": 9.23135738264467e-05, + "loss": 0.0215, + "step": 2200 + }, + { + "epoch": 7.342192691029901, + "grad_norm": 0.2535257339477539, + "learning_rate": 9.222525370090849e-05, + "loss": 0.0183, + "step": 2210 + }, + { + "epoch": 7.3754152823920265, + "grad_norm": 0.3889135718345642, + "learning_rate": 9.213647180676088e-05, + "loss": 0.0177, + "step": 2220 + }, + { + "epoch": 7.408637873754153, + "grad_norm": 0.3870062232017517, + "learning_rate": 9.204722911490846e-05, + "loss": 0.0226, + "step": 2230 + }, + { + "epoch": 7.441860465116279, + "grad_norm": 0.4052248001098633, + "learning_rate": 9.1957526601295e-05, + "loss": 0.0234, + "step": 2240 + }, + { + "epoch": 7.475083056478406, + "grad_norm": 0.28612571954727173, + "learning_rate": 9.186736524689281e-05, + "loss": 0.0196, + "step": 2250 + }, + { + "epoch": 7.5083056478405314, + "grad_norm": 0.35402119159698486, + "learning_rate": 9.177674603769204e-05, + "loss": 0.0242, + "step": 2260 + }, + { + "epoch": 7.541528239202658, + "grad_norm": 0.3981899619102478, + "learning_rate": 9.168566996468983e-05, + "loss": 0.0236, + "step": 2270 + }, + { + "epoch": 7.574750830564784, + "grad_norm": 0.20559972524642944, + "learning_rate": 9.159413802387951e-05, + "loss": 0.0228, + "step": 2280 + }, + { + "epoch": 7.607973421926911, + "grad_norm": 0.25476157665252686, + "learning_rate": 9.150215121623974e-05, + "loss": 0.0224, + "step": 2290 + }, + { + "epoch": 7.641196013289036, + "grad_norm": 0.22289226949214935, + "learning_rate": 9.140971054772349e-05, + "loss": 0.0213, + "step": 2300 + }, + { + "epoch": 7.674418604651163, + "grad_norm": 0.266042023897171, + "learning_rate": 9.131681702924713e-05, + "loss": 0.0221, + "step": 2310 + }, + { + "epoch": 7.707641196013289, + "grad_norm": 0.24307210743427277, + "learning_rate": 9.122347167667926e-05, + "loss": 0.021, + "step": 2320 + }, + { + "epoch": 7.740863787375416, + "grad_norm": 0.36476147174835205, + "learning_rate": 9.112967551082973e-05, + "loss": 0.0215, + "step": 2330 + }, + { + "epoch": 7.774086378737541, + "grad_norm": 0.2570967674255371, + "learning_rate": 9.103542955743835e-05, + "loss": 0.0264, + "step": 2340 + }, + { + "epoch": 7.807308970099668, + "grad_norm": 0.30510249733924866, + "learning_rate": 9.094073484716381e-05, + "loss": 0.023, + "step": 2350 + }, + { + "epoch": 7.840531561461794, + "grad_norm": 0.3401687741279602, + "learning_rate": 9.084559241557226e-05, + "loss": 0.0243, + "step": 2360 + }, + { + "epoch": 7.8737541528239205, + "grad_norm": 0.4495038390159607, + "learning_rate": 9.075000330312608e-05, + "loss": 0.0186, + "step": 2370 + }, + { + "epoch": 7.906976744186046, + "grad_norm": 0.22529259324073792, + "learning_rate": 9.065396855517253e-05, + "loss": 0.0197, + "step": 2380 + }, + { + "epoch": 7.940199335548173, + "grad_norm": 0.30005428194999695, + "learning_rate": 9.055748922193219e-05, + "loss": 0.0184, + "step": 2390 + }, + { + "epoch": 7.973421926910299, + "grad_norm": 0.3295758068561554, + "learning_rate": 9.046056635848761e-05, + "loss": 0.0231, + "step": 2400 + }, + { + "epoch": 8.006644518272426, + "grad_norm": 0.40123850107192993, + "learning_rate": 9.036320102477169e-05, + "loss": 0.02, + "step": 2410 + }, + { + "epoch": 8.039867109634551, + "grad_norm": 0.22583815455436707, + "learning_rate": 9.02653942855561e-05, + "loss": 0.0208, + "step": 2420 + }, + { + "epoch": 8.073089700996677, + "grad_norm": 0.3022814393043518, + "learning_rate": 9.016714721043971e-05, + "loss": 0.0193, + "step": 2430 + }, + { + "epoch": 8.106312292358805, + "grad_norm": 0.433930903673172, + "learning_rate": 9.006846087383675e-05, + "loss": 0.0233, + "step": 2440 + }, + { + "epoch": 8.13953488372093, + "grad_norm": 0.4505160450935364, + "learning_rate": 8.996933635496523e-05, + "loss": 0.0196, + "step": 2450 + }, + { + "epoch": 8.172757475083056, + "grad_norm": 0.4341692328453064, + "learning_rate": 8.986977473783498e-05, + "loss": 0.0216, + "step": 2460 + }, + { + "epoch": 8.205980066445182, + "grad_norm": 0.39989882707595825, + "learning_rate": 8.97697771112359e-05, + "loss": 0.0231, + "step": 2470 + }, + { + "epoch": 8.23920265780731, + "grad_norm": 0.31618350744247437, + "learning_rate": 8.966934456872602e-05, + "loss": 0.0159, + "step": 2480 + }, + { + "epoch": 8.272425249169435, + "grad_norm": 0.3098665475845337, + "learning_rate": 8.95684782086195e-05, + "loss": 0.0177, + "step": 2490 + }, + { + "epoch": 8.305647840531561, + "grad_norm": 0.24335923790931702, + "learning_rate": 8.946717913397476e-05, + "loss": 0.0191, + "step": 2500 + }, + { + "epoch": 8.338870431893687, + "grad_norm": 0.31290650367736816, + "learning_rate": 8.93654484525822e-05, + "loss": 0.0199, + "step": 2510 + }, + { + "epoch": 8.372093023255815, + "grad_norm": 0.4052436947822571, + "learning_rate": 8.926328727695226e-05, + "loss": 0.021, + "step": 2520 + }, + { + "epoch": 8.40531561461794, + "grad_norm": 0.432539701461792, + "learning_rate": 8.916069672430319e-05, + "loss": 0.0197, + "step": 2530 + }, + { + "epoch": 8.438538205980066, + "grad_norm": 0.30588874220848083, + "learning_rate": 8.905767791654884e-05, + "loss": 0.0155, + "step": 2540 + }, + { + "epoch": 8.471760797342192, + "grad_norm": 0.2926918566226959, + "learning_rate": 8.895423198028638e-05, + "loss": 0.0213, + "step": 2550 + }, + { + "epoch": 8.50498338870432, + "grad_norm": 0.2769851088523865, + "learning_rate": 8.885036004678402e-05, + "loss": 0.0247, + "step": 2560 + }, + { + "epoch": 8.538205980066445, + "grad_norm": 0.48992279171943665, + "learning_rate": 8.874606325196857e-05, + "loss": 0.0213, + "step": 2570 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.3409804105758667, + "learning_rate": 8.864134273641304e-05, + "loss": 0.0215, + "step": 2580 + }, + { + "epoch": 8.604651162790697, + "grad_norm": 0.2671858072280884, + "learning_rate": 8.853619964532427e-05, + "loss": 0.0189, + "step": 2590 + }, + { + "epoch": 8.637873754152825, + "grad_norm": 0.29365599155426025, + "learning_rate": 8.843063512853019e-05, + "loss": 0.0203, + "step": 2600 + }, + { + "epoch": 8.67109634551495, + "grad_norm": 0.35508570075035095, + "learning_rate": 8.832465034046749e-05, + "loss": 0.0198, + "step": 2610 + }, + { + "epoch": 8.704318936877076, + "grad_norm": 0.31511878967285156, + "learning_rate": 8.821824644016882e-05, + "loss": 0.0191, + "step": 2620 + }, + { + "epoch": 8.737541528239202, + "grad_norm": 0.2841794490814209, + "learning_rate": 8.811142459125019e-05, + "loss": 0.0184, + "step": 2630 + }, + { + "epoch": 8.77076411960133, + "grad_norm": 0.32950523495674133, + "learning_rate": 8.800418596189822e-05, + "loss": 0.0159, + "step": 2640 + }, + { + "epoch": 8.803986710963455, + "grad_norm": 0.23667795956134796, + "learning_rate": 8.789653172485737e-05, + "loss": 0.0148, + "step": 2650 + }, + { + "epoch": 8.837209302325581, + "grad_norm": 0.3478001058101654, + "learning_rate": 8.778846305741715e-05, + "loss": 0.0273, + "step": 2660 + }, + { + "epoch": 8.870431893687707, + "grad_norm": 0.4005625247955322, + "learning_rate": 8.767998114139918e-05, + "loss": 0.0207, + "step": 2670 + }, + { + "epoch": 8.903654485049834, + "grad_norm": 0.20847535133361816, + "learning_rate": 8.757108716314429e-05, + "loss": 0.0183, + "step": 2680 + }, + { + "epoch": 8.93687707641196, + "grad_norm": 0.1629849672317505, + "learning_rate": 8.746178231349962e-05, + "loss": 0.0234, + "step": 2690 + }, + { + "epoch": 8.970099667774086, + "grad_norm": 0.24203428626060486, + "learning_rate": 8.735206778780549e-05, + "loss": 0.0184, + "step": 2700 + }, + { + "epoch": 9.003322259136212, + "grad_norm": 0.2886161208152771, + "learning_rate": 8.724194478588234e-05, + "loss": 0.0193, + "step": 2710 + }, + { + "epoch": 9.03654485049834, + "grad_norm": 0.3591567575931549, + "learning_rate": 8.713141451201772e-05, + "loss": 0.0208, + "step": 2720 + }, + { + "epoch": 9.069767441860465, + "grad_norm": 0.3133392333984375, + "learning_rate": 8.702047817495295e-05, + "loss": 0.0195, + "step": 2730 + }, + { + "epoch": 9.102990033222591, + "grad_norm": 0.37037843465805054, + "learning_rate": 8.69091369878701e-05, + "loss": 0.019, + "step": 2740 + }, + { + "epoch": 9.136212624584717, + "grad_norm": 0.2838710844516754, + "learning_rate": 8.679739216837849e-05, + "loss": 0.0176, + "step": 2750 + }, + { + "epoch": 9.169435215946844, + "grad_norm": 0.41625136137008667, + "learning_rate": 8.66852449385016e-05, + "loss": 0.0176, + "step": 2760 + }, + { + "epoch": 9.20265780730897, + "grad_norm": 0.38990315794944763, + "learning_rate": 8.657269652466356e-05, + "loss": 0.0179, + "step": 2770 + }, + { + "epoch": 9.235880398671096, + "grad_norm": 0.2293078452348709, + "learning_rate": 8.645974815767577e-05, + "loss": 0.0173, + "step": 2780 + }, + { + "epoch": 9.269102990033222, + "grad_norm": 0.27936089038848877, + "learning_rate": 8.634640107272351e-05, + "loss": 0.0214, + "step": 2790 + }, + { + "epoch": 9.30232558139535, + "grad_norm": 0.3115111291408539, + "learning_rate": 8.623265650935234e-05, + "loss": 0.016, + "step": 2800 + }, + { + "epoch": 9.335548172757475, + "grad_norm": 0.4285588264465332, + "learning_rate": 8.611851571145456e-05, + "loss": 0.0213, + "step": 2810 + }, + { + "epoch": 9.368770764119601, + "grad_norm": 0.29907381534576416, + "learning_rate": 8.600397992725566e-05, + "loss": 0.0209, + "step": 2820 + }, + { + "epoch": 9.401993355481727, + "grad_norm": 0.25026509165763855, + "learning_rate": 8.588905040930061e-05, + "loss": 0.0197, + "step": 2830 + }, + { + "epoch": 9.435215946843854, + "grad_norm": 0.3209597170352936, + "learning_rate": 8.577372841444022e-05, + "loss": 0.0172, + "step": 2840 + }, + { + "epoch": 9.46843853820598, + "grad_norm": 0.19637836515903473, + "learning_rate": 8.565801520381736e-05, + "loss": 0.0192, + "step": 2850 + }, + { + "epoch": 9.501661129568106, + "grad_norm": 0.2835908532142639, + "learning_rate": 8.554191204285313e-05, + "loss": 0.0195, + "step": 2860 + }, + { + "epoch": 9.534883720930232, + "grad_norm": 0.3389662802219391, + "learning_rate": 8.542542020123315e-05, + "loss": 0.0173, + "step": 2870 + }, + { + "epoch": 9.56810631229236, + "grad_norm": 0.3397357165813446, + "learning_rate": 8.530854095289347e-05, + "loss": 0.019, + "step": 2880 + }, + { + "epoch": 9.601328903654485, + "grad_norm": 0.28326690196990967, + "learning_rate": 8.519127557600688e-05, + "loss": 0.0156, + "step": 2890 + }, + { + "epoch": 9.634551495016611, + "grad_norm": 0.43105173110961914, + "learning_rate": 8.507362535296871e-05, + "loss": 0.0169, + "step": 2900 + }, + { + "epoch": 9.667774086378738, + "grad_norm": 0.2692944407463074, + "learning_rate": 8.495559157038299e-05, + "loss": 0.0218, + "step": 2910 + }, + { + "epoch": 9.700996677740864, + "grad_norm": 0.21992726624011993, + "learning_rate": 8.483717551904823e-05, + "loss": 0.0259, + "step": 2920 + }, + { + "epoch": 9.73421926910299, + "grad_norm": 0.18732020258903503, + "learning_rate": 8.47183784939434e-05, + "loss": 0.0168, + "step": 2930 + }, + { + "epoch": 9.767441860465116, + "grad_norm": 0.38717469573020935, + "learning_rate": 8.459920179421374e-05, + "loss": 0.0183, + "step": 2940 + }, + { + "epoch": 9.800664451827242, + "grad_norm": 0.24515467882156372, + "learning_rate": 8.447964672315656e-05, + "loss": 0.0169, + "step": 2950 + }, + { + "epoch": 9.83388704318937, + "grad_norm": 0.25871196389198303, + "learning_rate": 8.435971458820692e-05, + "loss": 0.015, + "step": 2960 + }, + { + "epoch": 9.867109634551495, + "grad_norm": 0.19653911888599396, + "learning_rate": 8.423940670092345e-05, + "loss": 0.0195, + "step": 2970 + }, + { + "epoch": 9.90033222591362, + "grad_norm": 0.36540013551712036, + "learning_rate": 8.411872437697394e-05, + "loss": 0.0216, + "step": 2980 + }, + { + "epoch": 9.933554817275748, + "grad_norm": 0.21347872912883759, + "learning_rate": 8.399766893612096e-05, + "loss": 0.0168, + "step": 2990 + }, + { + "epoch": 9.966777408637874, + "grad_norm": 0.2640397548675537, + "learning_rate": 8.38762417022074e-05, + "loss": 0.0146, + "step": 3000 + }, + { + "epoch": 10.0, + "grad_norm": 0.9582907557487488, + "learning_rate": 8.375444400314204e-05, + "loss": 0.0201, + "step": 3010 + }, + { + "epoch": 10.033222591362126, + "grad_norm": 0.546707808971405, + "learning_rate": 8.3632277170885e-05, + "loss": 0.0231, + "step": 3020 + }, + { + "epoch": 10.066445182724253, + "grad_norm": 0.36035218834877014, + "learning_rate": 8.350974254143318e-05, + "loss": 0.0244, + "step": 3030 + }, + { + "epoch": 10.09966777408638, + "grad_norm": 0.40394413471221924, + "learning_rate": 8.338684145480566e-05, + "loss": 0.024, + "step": 3040 + }, + { + "epoch": 10.132890365448505, + "grad_norm": 0.3483884334564209, + "learning_rate": 8.326357525502904e-05, + "loss": 0.0199, + "step": 3050 + }, + { + "epoch": 10.16611295681063, + "grad_norm": 0.30480167269706726, + "learning_rate": 8.313994529012273e-05, + "loss": 0.023, + "step": 3060 + }, + { + "epoch": 10.199335548172758, + "grad_norm": 0.29263198375701904, + "learning_rate": 8.301595291208422e-05, + "loss": 0.0194, + "step": 3070 + }, + { + "epoch": 10.232558139534884, + "grad_norm": 0.37630799412727356, + "learning_rate": 8.289159947687427e-05, + "loss": 0.0227, + "step": 3080 + }, + { + "epoch": 10.26578073089701, + "grad_norm": 0.316376268863678, + "learning_rate": 8.276688634440216e-05, + "loss": 0.0176, + "step": 3090 + }, + { + "epoch": 10.299003322259136, + "grad_norm": 0.28122377395629883, + "learning_rate": 8.26418148785107e-05, + "loss": 0.0219, + "step": 3100 + }, + { + "epoch": 10.332225913621263, + "grad_norm": 0.36189690232276917, + "learning_rate": 8.251638644696141e-05, + "loss": 0.0196, + "step": 3110 + }, + { + "epoch": 10.365448504983389, + "grad_norm": 0.31648537516593933, + "learning_rate": 8.23906024214195e-05, + "loss": 0.0227, + "step": 3120 + }, + { + "epoch": 10.398671096345515, + "grad_norm": 0.36495092511177063, + "learning_rate": 8.226446417743897e-05, + "loss": 0.0156, + "step": 3130 + }, + { + "epoch": 10.43189368770764, + "grad_norm": 0.3638916313648224, + "learning_rate": 8.213797309444742e-05, + "loss": 0.0228, + "step": 3140 + }, + { + "epoch": 10.465116279069768, + "grad_norm": 0.2252400666475296, + "learning_rate": 8.201113055573105e-05, + "loss": 0.0195, + "step": 3150 + }, + { + "epoch": 10.498338870431894, + "grad_norm": 0.5001240372657776, + "learning_rate": 8.188393794841958e-05, + "loss": 0.0216, + "step": 3160 + }, + { + "epoch": 10.53156146179402, + "grad_norm": 0.3034009635448456, + "learning_rate": 8.175639666347094e-05, + "loss": 0.0192, + "step": 3170 + }, + { + "epoch": 10.564784053156146, + "grad_norm": 0.32421091198921204, + "learning_rate": 8.162850809565623e-05, + "loss": 0.0193, + "step": 3180 + }, + { + "epoch": 10.598006644518273, + "grad_norm": 0.2746524214744568, + "learning_rate": 8.150027364354431e-05, + "loss": 0.0176, + "step": 3190 + }, + { + "epoch": 10.631229235880399, + "grad_norm": 0.5117692351341248, + "learning_rate": 8.137169470948662e-05, + "loss": 0.0217, + "step": 3200 + }, + { + "epoch": 10.664451827242525, + "grad_norm": 0.322994589805603, + "learning_rate": 8.124277269960179e-05, + "loss": 0.0187, + "step": 3210 + }, + { + "epoch": 10.69767441860465, + "grad_norm": 0.18638882040977478, + "learning_rate": 8.111350902376023e-05, + "loss": 0.0173, + "step": 3220 + }, + { + "epoch": 10.730897009966778, + "grad_norm": 0.2154085487127304, + "learning_rate": 8.098390509556883e-05, + "loss": 0.0151, + "step": 3230 + }, + { + "epoch": 10.764119601328904, + "grad_norm": 0.14285944402217865, + "learning_rate": 8.085396233235536e-05, + "loss": 0.018, + "step": 3240 + }, + { + "epoch": 10.79734219269103, + "grad_norm": 0.34248557686805725, + "learning_rate": 8.072368215515306e-05, + "loss": 0.0146, + "step": 3250 + }, + { + "epoch": 10.830564784053156, + "grad_norm": 0.2130262702703476, + "learning_rate": 8.059306598868506e-05, + "loss": 0.0124, + "step": 3260 + }, + { + "epoch": 10.863787375415283, + "grad_norm": 0.40157315135002136, + "learning_rate": 8.046211526134888e-05, + "loss": 0.0257, + "step": 3270 + }, + { + "epoch": 10.897009966777409, + "grad_norm": 0.38071388006210327, + "learning_rate": 8.033083140520065e-05, + "loss": 0.0186, + "step": 3280 + }, + { + "epoch": 10.930232558139535, + "grad_norm": 0.24200329184532166, + "learning_rate": 8.019921585593962e-05, + "loss": 0.0161, + "step": 3290 + }, + { + "epoch": 10.96345514950166, + "grad_norm": 0.23784863948822021, + "learning_rate": 8.006727005289232e-05, + "loss": 0.0169, + "step": 3300 + }, + { + "epoch": 10.996677740863788, + "grad_norm": 0.2810799777507782, + "learning_rate": 7.993499543899692e-05, + "loss": 0.0175, + "step": 3310 + }, + { + "epoch": 11.029900332225914, + "grad_norm": 0.3434373140335083, + "learning_rate": 7.980239346078742e-05, + "loss": 0.0238, + "step": 3320 + }, + { + "epoch": 11.06312292358804, + "grad_norm": 0.4956740438938141, + "learning_rate": 7.966946556837778e-05, + "loss": 0.0219, + "step": 3330 + }, + { + "epoch": 11.096345514950166, + "grad_norm": 0.22429168224334717, + "learning_rate": 7.953621321544616e-05, + "loss": 0.0221, + "step": 3340 + }, + { + "epoch": 11.129568106312293, + "grad_norm": 0.34100890159606934, + "learning_rate": 7.940263785921896e-05, + "loss": 0.0217, + "step": 3350 + }, + { + "epoch": 11.162790697674419, + "grad_norm": 0.3702620565891266, + "learning_rate": 7.926874096045482e-05, + "loss": 0.028, + "step": 3360 + }, + { + "epoch": 11.196013289036545, + "grad_norm": 0.3129262626171112, + "learning_rate": 7.913452398342881e-05, + "loss": 0.018, + "step": 3370 + }, + { + "epoch": 11.22923588039867, + "grad_norm": 0.3758520483970642, + "learning_rate": 7.89999883959163e-05, + "loss": 0.0253, + "step": 3380 + }, + { + "epoch": 11.262458471760798, + "grad_norm": 0.36128005385398865, + "learning_rate": 7.886513566917687e-05, + "loss": 0.0231, + "step": 3390 + }, + { + "epoch": 11.295681063122924, + "grad_norm": 0.3874170482158661, + "learning_rate": 7.872996727793838e-05, + "loss": 0.0193, + "step": 3400 + }, + { + "epoch": 11.32890365448505, + "grad_norm": 0.4195541739463806, + "learning_rate": 7.859448470038069e-05, + "loss": 0.0202, + "step": 3410 + }, + { + "epoch": 11.362126245847175, + "grad_norm": 0.36057302355766296, + "learning_rate": 7.845868941811956e-05, + "loss": 0.0216, + "step": 3420 + }, + { + "epoch": 11.395348837209303, + "grad_norm": 0.3191004991531372, + "learning_rate": 7.832258291619043e-05, + "loss": 0.0207, + "step": 3430 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.3054732084274292, + "learning_rate": 7.81861666830322e-05, + "loss": 0.0175, + "step": 3440 + }, + { + "epoch": 11.461794019933555, + "grad_norm": 0.2970241606235504, + "learning_rate": 7.804944221047097e-05, + "loss": 0.0213, + "step": 3450 + }, + { + "epoch": 11.49501661129568, + "grad_norm": 0.3414061367511749, + "learning_rate": 7.791241099370364e-05, + "loss": 0.0177, + "step": 3460 + }, + { + "epoch": 11.528239202657808, + "grad_norm": 0.4683157205581665, + "learning_rate": 7.777507453128163e-05, + "loss": 0.0242, + "step": 3470 + }, + { + "epoch": 11.561461794019934, + "grad_norm": 0.21441063284873962, + "learning_rate": 7.763743432509451e-05, + "loss": 0.0169, + "step": 3480 + }, + { + "epoch": 11.59468438538206, + "grad_norm": 0.29838085174560547, + "learning_rate": 7.749949188035353e-05, + "loss": 0.0188, + "step": 3490 + }, + { + "epoch": 11.627906976744185, + "grad_norm": 0.3362698554992676, + "learning_rate": 7.736124870557516e-05, + "loss": 0.019, + "step": 3500 + }, + { + "epoch": 11.661129568106313, + "grad_norm": 0.27162593603134155, + "learning_rate": 7.722270631256459e-05, + "loss": 0.0257, + "step": 3510 + }, + { + "epoch": 11.694352159468439, + "grad_norm": 0.2674727737903595, + "learning_rate": 7.708386621639925e-05, + "loss": 0.0201, + "step": 3520 + }, + { + "epoch": 11.727574750830565, + "grad_norm": 0.25655072927474976, + "learning_rate": 7.694472993541219e-05, + "loss": 0.0179, + "step": 3530 + }, + { + "epoch": 11.76079734219269, + "grad_norm": 0.24716055393218994, + "learning_rate": 7.680529899117547e-05, + "loss": 0.0181, + "step": 3540 + }, + { + "epoch": 11.794019933554818, + "grad_norm": 0.4514801800251007, + "learning_rate": 7.666557490848358e-05, + "loss": 0.0247, + "step": 3550 + }, + { + "epoch": 11.827242524916944, + "grad_norm": 0.5603443384170532, + "learning_rate": 7.65255592153367e-05, + "loss": 0.0217, + "step": 3560 + }, + { + "epoch": 11.86046511627907, + "grad_norm": 0.32442739605903625, + "learning_rate": 7.638525344292402e-05, + "loss": 0.0214, + "step": 3570 + }, + { + "epoch": 11.893687707641195, + "grad_norm": 0.27662280201911926, + "learning_rate": 7.624465912560697e-05, + "loss": 0.0221, + "step": 3580 + }, + { + "epoch": 11.926910299003323, + "grad_norm": 0.22196006774902344, + "learning_rate": 7.610377780090249e-05, + "loss": 0.0214, + "step": 3590 + }, + { + "epoch": 11.960132890365449, + "grad_norm": 0.36716607213020325, + "learning_rate": 7.596261100946618e-05, + "loss": 0.0182, + "step": 3600 + }, + { + "epoch": 11.993355481727574, + "grad_norm": 0.23657159507274628, + "learning_rate": 7.582116029507542e-05, + "loss": 0.0228, + "step": 3610 + }, + { + "epoch": 12.0265780730897, + "grad_norm": 0.3037751317024231, + "learning_rate": 7.56794272046126e-05, + "loss": 0.0158, + "step": 3620 + }, + { + "epoch": 12.059800664451828, + "grad_norm": 0.21974584460258484, + "learning_rate": 7.55374132880481e-05, + "loss": 0.0181, + "step": 3630 + }, + { + "epoch": 12.093023255813954, + "grad_norm": 0.20289649069309235, + "learning_rate": 7.539512009842333e-05, + "loss": 0.0179, + "step": 3640 + }, + { + "epoch": 12.12624584717608, + "grad_norm": 0.2716876268386841, + "learning_rate": 7.525254919183382e-05, + "loss": 0.0213, + "step": 3650 + }, + { + "epoch": 12.159468438538205, + "grad_norm": 0.2261294424533844, + "learning_rate": 7.510970212741215e-05, + "loss": 0.0198, + "step": 3660 + }, + { + "epoch": 12.192691029900333, + "grad_norm": 0.25505638122558594, + "learning_rate": 7.496658046731096e-05, + "loss": 0.0167, + "step": 3670 + }, + { + "epoch": 12.225913621262459, + "grad_norm": 0.23943322896957397, + "learning_rate": 7.482318577668578e-05, + "loss": 0.0176, + "step": 3680 + }, + { + "epoch": 12.259136212624584, + "grad_norm": 0.4076962172985077, + "learning_rate": 7.467951962367796e-05, + "loss": 0.0186, + "step": 3690 + }, + { + "epoch": 12.29235880398671, + "grad_norm": 0.34784767031669617, + "learning_rate": 7.453558357939755e-05, + "loss": 0.0212, + "step": 3700 + }, + { + "epoch": 12.325581395348838, + "grad_norm": 0.342916876077652, + "learning_rate": 7.439137921790606e-05, + "loss": 0.0163, + "step": 3710 + }, + { + "epoch": 12.358803986710964, + "grad_norm": 0.3107483386993408, + "learning_rate": 7.42469081161993e-05, + "loss": 0.0206, + "step": 3720 + }, + { + "epoch": 12.39202657807309, + "grad_norm": 0.3539960980415344, + "learning_rate": 7.410217185419006e-05, + "loss": 0.0174, + "step": 3730 + }, + { + "epoch": 12.425249169435215, + "grad_norm": 0.3659731149673462, + "learning_rate": 7.395717201469095e-05, + "loss": 0.0167, + "step": 3740 + }, + { + "epoch": 12.458471760797343, + "grad_norm": 0.2727093994617462, + "learning_rate": 7.381191018339696e-05, + "loss": 0.019, + "step": 3750 + }, + { + "epoch": 12.491694352159469, + "grad_norm": 0.41176939010620117, + "learning_rate": 7.36663879488682e-05, + "loss": 0.0176, + "step": 3760 + }, + { + "epoch": 12.524916943521594, + "grad_norm": 0.4040158987045288, + "learning_rate": 7.352060690251254e-05, + "loss": 0.02, + "step": 3770 + }, + { + "epoch": 12.55813953488372, + "grad_norm": 0.3903579115867615, + "learning_rate": 7.337456863856811e-05, + "loss": 0.0183, + "step": 3780 + }, + { + "epoch": 12.591362126245848, + "grad_norm": 0.3343384265899658, + "learning_rate": 7.3228274754086e-05, + "loss": 0.0227, + "step": 3790 + }, + { + "epoch": 12.624584717607974, + "grad_norm": 0.35453101992607117, + "learning_rate": 7.308172684891267e-05, + "loss": 0.0146, + "step": 3800 + }, + { + "epoch": 12.6578073089701, + "grad_norm": 0.35144782066345215, + "learning_rate": 7.293492652567255e-05, + "loss": 0.0196, + "step": 3810 + }, + { + "epoch": 12.691029900332225, + "grad_norm": 0.35027822852134705, + "learning_rate": 7.278787538975043e-05, + "loss": 0.0149, + "step": 3820 + }, + { + "epoch": 12.724252491694353, + "grad_norm": 0.34613221883773804, + "learning_rate": 7.2640575049274e-05, + "loss": 0.0207, + "step": 3830 + }, + { + "epoch": 12.757475083056478, + "grad_norm": 0.34608450531959534, + "learning_rate": 7.249302711509616e-05, + "loss": 0.0173, + "step": 3840 + }, + { + "epoch": 12.790697674418604, + "grad_norm": 0.31654372811317444, + "learning_rate": 7.23452332007775e-05, + "loss": 0.0146, + "step": 3850 + }, + { + "epoch": 12.82392026578073, + "grad_norm": 0.2791317403316498, + "learning_rate": 7.219719492256858e-05, + "loss": 0.015, + "step": 3860 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.44848188757896423, + "learning_rate": 7.20489138993923e-05, + "loss": 0.0175, + "step": 3870 + }, + { + "epoch": 12.890365448504983, + "grad_norm": 0.3050365149974823, + "learning_rate": 7.190039175282614e-05, + "loss": 0.0174, + "step": 3880 + }, + { + "epoch": 12.92358803986711, + "grad_norm": 0.290131539106369, + "learning_rate": 7.175163010708455e-05, + "loss": 0.0194, + "step": 3890 + }, + { + "epoch": 12.956810631229235, + "grad_norm": 0.3899692893028259, + "learning_rate": 7.1602630589001e-05, + "loss": 0.0141, + "step": 3900 + }, + { + "epoch": 12.990033222591363, + "grad_norm": 0.25462275743484497, + "learning_rate": 7.14533948280104e-05, + "loss": 0.021, + "step": 3910 + }, + { + "epoch": 13.023255813953488, + "grad_norm": 0.25204241275787354, + "learning_rate": 7.130392445613109e-05, + "loss": 0.0159, + "step": 3920 + }, + { + "epoch": 13.056478405315614, + "grad_norm": 0.35123640298843384, + "learning_rate": 7.115422110794711e-05, + "loss": 0.0189, + "step": 3930 + }, + { + "epoch": 13.08970099667774, + "grad_norm": 0.3995964229106903, + "learning_rate": 7.100428642059033e-05, + "loss": 0.0175, + "step": 3940 + }, + { + "epoch": 13.122923588039868, + "grad_norm": 0.3554200530052185, + "learning_rate": 7.08541220337224e-05, + "loss": 0.0191, + "step": 3950 + }, + { + "epoch": 13.156146179401993, + "grad_norm": 0.3538668751716614, + "learning_rate": 7.070372958951706e-05, + "loss": 0.0186, + "step": 3960 + }, + { + "epoch": 13.18936877076412, + "grad_norm": 0.1743728220462799, + "learning_rate": 7.055311073264194e-05, + "loss": 0.016, + "step": 3970 + }, + { + "epoch": 13.222591362126245, + "grad_norm": 0.287780225276947, + "learning_rate": 7.040226711024077e-05, + "loss": 0.0187, + "step": 3980 + }, + { + "epoch": 13.255813953488373, + "grad_norm": 0.24803023040294647, + "learning_rate": 7.02512003719152e-05, + "loss": 0.0179, + "step": 3990 + }, + { + "epoch": 13.289036544850498, + "grad_norm": 0.2810683846473694, + "learning_rate": 7.00999121697069e-05, + "loss": 0.0181, + "step": 4000 + }, + { + "epoch": 13.322259136212624, + "grad_norm": 0.31502121686935425, + "learning_rate": 6.99484041580794e-05, + "loss": 0.0205, + "step": 4010 + }, + { + "epoch": 13.355481727574752, + "grad_norm": 0.23561535775661469, + "learning_rate": 6.979667799390004e-05, + "loss": 0.0174, + "step": 4020 + }, + { + "epoch": 13.388704318936878, + "grad_norm": 0.36243340373039246, + "learning_rate": 6.964473533642185e-05, + "loss": 0.014, + "step": 4030 + }, + { + "epoch": 13.421926910299003, + "grad_norm": 0.24262340366840363, + "learning_rate": 6.949257784726539e-05, + "loss": 0.0171, + "step": 4040 + }, + { + "epoch": 13.455149501661129, + "grad_norm": 0.23523877561092377, + "learning_rate": 6.934020719040056e-05, + "loss": 0.0173, + "step": 4050 + }, + { + "epoch": 13.488372093023255, + "grad_norm": 0.2318788468837738, + "learning_rate": 6.918762503212848e-05, + "loss": 0.0156, + "step": 4060 + }, + { + "epoch": 13.521594684385382, + "grad_norm": 0.265948623418808, + "learning_rate": 6.903483304106319e-05, + "loss": 0.0169, + "step": 4070 + }, + { + "epoch": 13.554817275747508, + "grad_norm": 0.2921631932258606, + "learning_rate": 6.888183288811341e-05, + "loss": 0.0173, + "step": 4080 + }, + { + "epoch": 13.588039867109634, + "grad_norm": 0.2878166735172272, + "learning_rate": 6.87286262464643e-05, + "loss": 0.014, + "step": 4090 + }, + { + "epoch": 13.621262458471762, + "grad_norm": 0.36211255192756653, + "learning_rate": 6.857521479155915e-05, + "loss": 0.0124, + "step": 4100 + }, + { + "epoch": 13.654485049833887, + "grad_norm": 0.4308359920978546, + "learning_rate": 6.842160020108104e-05, + "loss": 0.0184, + "step": 4110 + }, + { + "epoch": 13.687707641196013, + "grad_norm": 0.3886549174785614, + "learning_rate": 6.826778415493455e-05, + "loss": 0.0175, + "step": 4120 + }, + { + "epoch": 13.720930232558139, + "grad_norm": 0.29117047786712646, + "learning_rate": 6.811376833522729e-05, + "loss": 0.0182, + "step": 4130 + }, + { + "epoch": 13.754152823920267, + "grad_norm": 0.26800787448883057, + "learning_rate": 6.795955442625159e-05, + "loss": 0.0146, + "step": 4140 + }, + { + "epoch": 13.787375415282392, + "grad_norm": 0.30751490592956543, + "learning_rate": 6.780514411446608e-05, + "loss": 0.0187, + "step": 4150 + }, + { + "epoch": 13.820598006644518, + "grad_norm": 0.36831483244895935, + "learning_rate": 6.765053908847716e-05, + "loss": 0.0161, + "step": 4160 + }, + { + "epoch": 13.853820598006644, + "grad_norm": 0.32650047540664673, + "learning_rate": 6.749574103902064e-05, + "loss": 0.0114, + "step": 4170 + }, + { + "epoch": 13.887043189368772, + "grad_norm": 0.22459985315799713, + "learning_rate": 6.734075165894317e-05, + "loss": 0.0195, + "step": 4180 + }, + { + "epoch": 13.920265780730897, + "grad_norm": 0.2692617177963257, + "learning_rate": 6.71855726431838e-05, + "loss": 0.0178, + "step": 4190 + }, + { + "epoch": 13.953488372093023, + "grad_norm": 0.28635933995246887, + "learning_rate": 6.703020568875538e-05, + "loss": 0.0234, + "step": 4200 + }, + { + "epoch": 13.986710963455149, + "grad_norm": 0.1947304904460907, + "learning_rate": 6.687465249472603e-05, + "loss": 0.0137, + "step": 4210 + }, + { + "epoch": 14.019933554817277, + "grad_norm": 0.454317182302475, + "learning_rate": 6.671891476220055e-05, + "loss": 0.0205, + "step": 4220 + }, + { + "epoch": 14.053156146179402, + "grad_norm": 0.2946520447731018, + "learning_rate": 6.656299419430183e-05, + "loss": 0.0196, + "step": 4230 + }, + { + "epoch": 14.086378737541528, + "grad_norm": 0.3139476478099823, + "learning_rate": 6.640689249615223e-05, + "loss": 0.0153, + "step": 4240 + }, + { + "epoch": 14.119601328903654, + "grad_norm": 0.44135165214538574, + "learning_rate": 6.625061137485491e-05, + "loss": 0.0174, + "step": 4250 + }, + { + "epoch": 14.152823920265782, + "grad_norm": 0.3966526687145233, + "learning_rate": 6.609415253947517e-05, + "loss": 0.0181, + "step": 4260 + }, + { + "epoch": 14.186046511627907, + "grad_norm": 0.48178598284721375, + "learning_rate": 6.593751770102178e-05, + "loss": 0.0229, + "step": 4270 + }, + { + "epoch": 14.219269102990033, + "grad_norm": 0.4271031320095062, + "learning_rate": 6.578070857242823e-05, + "loss": 0.0173, + "step": 4280 + }, + { + "epoch": 14.252491694352159, + "grad_norm": 0.1896420121192932, + "learning_rate": 6.562372686853402e-05, + "loss": 0.0141, + "step": 4290 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.4319266974925995, + "learning_rate": 6.546657430606593e-05, + "loss": 0.0164, + "step": 4300 + }, + { + "epoch": 14.318936877076412, + "grad_norm": 0.24711720645427704, + "learning_rate": 6.530925260361918e-05, + "loss": 0.0161, + "step": 4310 + }, + { + "epoch": 14.352159468438538, + "grad_norm": 0.2857001721858978, + "learning_rate": 6.515176348163871e-05, + "loss": 0.014, + "step": 4320 + }, + { + "epoch": 14.385382059800664, + "grad_norm": 0.3837076723575592, + "learning_rate": 6.499410866240032e-05, + "loss": 0.0162, + "step": 4330 + }, + { + "epoch": 14.418604651162791, + "grad_norm": 0.23993246257305145, + "learning_rate": 6.48362898699919e-05, + "loss": 0.0138, + "step": 4340 + }, + { + "epoch": 14.451827242524917, + "grad_norm": 0.30589741468429565, + "learning_rate": 6.467830883029443e-05, + "loss": 0.0134, + "step": 4350 + }, + { + "epoch": 14.485049833887043, + "grad_norm": 0.26103895902633667, + "learning_rate": 6.452016727096326e-05, + "loss": 0.0147, + "step": 4360 + }, + { + "epoch": 14.518272425249169, + "grad_norm": 0.1985149085521698, + "learning_rate": 6.436186692140916e-05, + "loss": 0.0143, + "step": 4370 + }, + { + "epoch": 14.551495016611296, + "grad_norm": 0.2402988076210022, + "learning_rate": 6.420340951277938e-05, + "loss": 0.0127, + "step": 4380 + }, + { + "epoch": 14.584717607973422, + "grad_norm": 0.2162981629371643, + "learning_rate": 6.404479677793874e-05, + "loss": 0.0123, + "step": 4390 + }, + { + "epoch": 14.617940199335548, + "grad_norm": 0.21887542307376862, + "learning_rate": 6.388603045145075e-05, + "loss": 0.0144, + "step": 4400 + }, + { + "epoch": 14.651162790697674, + "grad_norm": 0.26267585158348083, + "learning_rate": 6.372711226955843e-05, + "loss": 0.0212, + "step": 4410 + }, + { + "epoch": 14.684385382059801, + "grad_norm": 0.3892922103404999, + "learning_rate": 6.356804397016564e-05, + "loss": 0.0181, + "step": 4420 + }, + { + "epoch": 14.717607973421927, + "grad_norm": 0.30290210247039795, + "learning_rate": 6.340882729281779e-05, + "loss": 0.0183, + "step": 4430 + }, + { + "epoch": 14.750830564784053, + "grad_norm": 0.2289399802684784, + "learning_rate": 6.324946397868294e-05, + "loss": 0.0177, + "step": 4440 + }, + { + "epoch": 14.784053156146179, + "grad_norm": 0.22953011095523834, + "learning_rate": 6.308995577053276e-05, + "loss": 0.0159, + "step": 4450 + }, + { + "epoch": 14.817275747508306, + "grad_norm": 0.2985853850841522, + "learning_rate": 6.293030441272347e-05, + "loss": 0.0206, + "step": 4460 + }, + { + "epoch": 14.850498338870432, + "grad_norm": 0.19090281426906586, + "learning_rate": 6.277051165117677e-05, + "loss": 0.0148, + "step": 4470 + }, + { + "epoch": 14.883720930232558, + "grad_norm": 0.17965582013130188, + "learning_rate": 6.261057923336064e-05, + "loss": 0.0149, + "step": 4480 + }, + { + "epoch": 14.916943521594684, + "grad_norm": 0.22538508474826813, + "learning_rate": 6.245050890827042e-05, + "loss": 0.0159, + "step": 4490 + }, + { + "epoch": 14.950166112956811, + "grad_norm": 0.2691894769668579, + "learning_rate": 6.229030242640952e-05, + "loss": 0.0172, + "step": 4500 + }, + { + "epoch": 14.983388704318937, + "grad_norm": 0.2683984637260437, + "learning_rate": 6.212996153977037e-05, + "loss": 0.0153, + "step": 4510 + }, + { + "epoch": 15.016611295681063, + "grad_norm": 0.3912666440010071, + "learning_rate": 6.196948800181523e-05, + "loss": 0.019, + "step": 4520 + }, + { + "epoch": 15.049833887043189, + "grad_norm": 0.22746188938617706, + "learning_rate": 6.180888356745695e-05, + "loss": 0.0222, + "step": 4530 + }, + { + "epoch": 15.083056478405316, + "grad_norm": 0.20105905830860138, + "learning_rate": 6.164814999303995e-05, + "loss": 0.0154, + "step": 4540 + }, + { + "epoch": 15.116279069767442, + "grad_norm": 0.24199578166007996, + "learning_rate": 6.148728903632081e-05, + "loss": 0.0166, + "step": 4550 + }, + { + "epoch": 15.149501661129568, + "grad_norm": 0.3318256139755249, + "learning_rate": 6.132630245644921e-05, + "loss": 0.0164, + "step": 4560 + }, + { + "epoch": 15.182724252491694, + "grad_norm": 0.41996484994888306, + "learning_rate": 6.116519201394857e-05, + "loss": 0.014, + "step": 4570 + }, + { + "epoch": 15.215946843853821, + "grad_norm": 0.22289037704467773, + "learning_rate": 6.10039594706969e-05, + "loss": 0.0158, + "step": 4580 + }, + { + "epoch": 15.249169435215947, + "grad_norm": 0.3248358368873596, + "learning_rate": 6.084260658990744e-05, + "loss": 0.0131, + "step": 4590 + }, + { + "epoch": 15.282392026578073, + "grad_norm": 0.3744429051876068, + "learning_rate": 6.068113513610943e-05, + "loss": 0.0161, + "step": 4600 + }, + { + "epoch": 15.315614617940199, + "grad_norm": 0.21573661267757416, + "learning_rate": 6.0519546875128876e-05, + "loss": 0.0145, + "step": 4610 + }, + { + "epoch": 15.348837209302326, + "grad_norm": 0.22471874952316284, + "learning_rate": 6.035784357406906e-05, + "loss": 0.0151, + "step": 4620 + }, + { + "epoch": 15.382059800664452, + "grad_norm": 0.2581987679004669, + "learning_rate": 6.01960270012914e-05, + "loss": 0.0141, + "step": 4630 + }, + { + "epoch": 15.415282392026578, + "grad_norm": 0.4127011001110077, + "learning_rate": 6.003409892639599e-05, + "loss": 0.0165, + "step": 4640 + }, + { + "epoch": 15.448504983388704, + "grad_norm": 0.21382664144039154, + "learning_rate": 5.9872061120202336e-05, + "loss": 0.0129, + "step": 4650 + }, + { + "epoch": 15.481727574750831, + "grad_norm": 0.33771613240242004, + "learning_rate": 5.9709915354729914e-05, + "loss": 0.0168, + "step": 4660 + }, + { + "epoch": 15.514950166112957, + "grad_norm": 0.22272400557994843, + "learning_rate": 5.9547663403178824e-05, + "loss": 0.0183, + "step": 4670 + }, + { + "epoch": 15.548172757475083, + "grad_norm": 0.2086458057165146, + "learning_rate": 5.9385307039910445e-05, + "loss": 0.0148, + "step": 4680 + }, + { + "epoch": 15.581395348837209, + "grad_norm": 0.3473312556743622, + "learning_rate": 5.922284804042792e-05, + "loss": 0.0145, + "step": 4690 + }, + { + "epoch": 15.614617940199336, + "grad_norm": 0.24149948358535767, + "learning_rate": 5.906028818135687e-05, + "loss": 0.017, + "step": 4700 + }, + { + "epoch": 15.647840531561462, + "grad_norm": 0.27326861023902893, + "learning_rate": 5.889762924042585e-05, + "loss": 0.0166, + "step": 4710 + }, + { + "epoch": 15.681063122923588, + "grad_norm": 0.2840353846549988, + "learning_rate": 5.873487299644699e-05, + "loss": 0.0164, + "step": 4720 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.20497111976146698, + "learning_rate": 5.857202122929649e-05, + "loss": 0.015, + "step": 4730 + }, + { + "epoch": 15.747508305647841, + "grad_norm": 0.35892006754875183, + "learning_rate": 5.840907571989518e-05, + "loss": 0.0178, + "step": 4740 + }, + { + "epoch": 15.780730897009967, + "grad_norm": 0.23509638011455536, + "learning_rate": 5.824603825018904e-05, + "loss": 0.0156, + "step": 4750 + }, + { + "epoch": 15.813953488372093, + "grad_norm": 0.36275961995124817, + "learning_rate": 5.808291060312975e-05, + "loss": 0.0147, + "step": 4760 + }, + { + "epoch": 15.847176079734218, + "grad_norm": 0.18457593023777008, + "learning_rate": 5.7919694562655083e-05, + "loss": 0.0179, + "step": 4770 + }, + { + "epoch": 15.880398671096346, + "grad_norm": 0.2808282971382141, + "learning_rate": 5.775639191366954e-05, + "loss": 0.0152, + "step": 4780 + }, + { + "epoch": 15.913621262458472, + "grad_norm": 0.3134106695652008, + "learning_rate": 5.75930044420247e-05, + "loss": 0.0176, + "step": 4790 + }, + { + "epoch": 15.946843853820598, + "grad_norm": 0.1334390789270401, + "learning_rate": 5.74295339344998e-05, + "loss": 0.0136, + "step": 4800 + }, + { + "epoch": 15.980066445182723, + "grad_norm": 0.1954159289598465, + "learning_rate": 5.726598217878211e-05, + "loss": 0.0127, + "step": 4810 + }, + { + "epoch": 16.01328903654485, + "grad_norm": 0.3972402513027191, + "learning_rate": 5.71023509634474e-05, + "loss": 0.0165, + "step": 4820 + }, + { + "epoch": 16.046511627906977, + "grad_norm": 0.2406802922487259, + "learning_rate": 5.693864207794049e-05, + "loss": 0.0191, + "step": 4830 + }, + { + "epoch": 16.079734219269103, + "grad_norm": 0.2822772264480591, + "learning_rate": 5.677485731255545e-05, + "loss": 0.0151, + "step": 4840 + }, + { + "epoch": 16.11295681063123, + "grad_norm": 0.3374737799167633, + "learning_rate": 5.6610998458416296e-05, + "loss": 0.0251, + "step": 4850 + }, + { + "epoch": 16.146179401993354, + "grad_norm": 0.23243625462055206, + "learning_rate": 5.644706730745716e-05, + "loss": 0.0172, + "step": 4860 + }, + { + "epoch": 16.17940199335548, + "grad_norm": 0.17415964603424072, + "learning_rate": 5.628306565240287e-05, + "loss": 0.0129, + "step": 4870 + }, + { + "epoch": 16.21262458471761, + "grad_norm": 0.24170176684856415, + "learning_rate": 5.611899528674923e-05, + "loss": 0.0167, + "step": 4880 + }, + { + "epoch": 16.245847176079735, + "grad_norm": 0.18468914926052094, + "learning_rate": 5.595485800474349e-05, + "loss": 0.0176, + "step": 4890 + }, + { + "epoch": 16.27906976744186, + "grad_norm": 0.17150849103927612, + "learning_rate": 5.579065560136467e-05, + "loss": 0.0126, + "step": 4900 + }, + { + "epoch": 16.312292358803987, + "grad_norm": 0.30511826276779175, + "learning_rate": 5.562638987230392e-05, + "loss": 0.0172, + "step": 4910 + }, + { + "epoch": 16.345514950166113, + "grad_norm": 0.27298447489738464, + "learning_rate": 5.546206261394498e-05, + "loss": 0.0154, + "step": 4920 + }, + { + "epoch": 16.37873754152824, + "grad_norm": 0.1913800686597824, + "learning_rate": 5.529767562334437e-05, + "loss": 0.0133, + "step": 4930 + }, + { + "epoch": 16.411960132890364, + "grad_norm": 0.22306601703166962, + "learning_rate": 5.5133230698211926e-05, + "loss": 0.0128, + "step": 4940 + }, + { + "epoch": 16.44518272425249, + "grad_norm": 0.29052093625068665, + "learning_rate": 5.496872963689096e-05, + "loss": 0.0148, + "step": 4950 + }, + { + "epoch": 16.47840531561462, + "grad_norm": 0.22746320068836212, + "learning_rate": 5.4804174238338756e-05, + "loss": 0.0171, + "step": 4960 + }, + { + "epoch": 16.511627906976745, + "grad_norm": 0.35950127243995667, + "learning_rate": 5.463956630210678e-05, + "loss": 0.0164, + "step": 4970 + }, + { + "epoch": 16.54485049833887, + "grad_norm": 0.3040778934955597, + "learning_rate": 5.4474907628321046e-05, + "loss": 0.013, + "step": 4980 + }, + { + "epoch": 16.578073089700997, + "grad_norm": 0.3150460124015808, + "learning_rate": 5.431020001766244e-05, + "loss": 0.0182, + "step": 4990 + }, + { + "epoch": 16.611295681063122, + "grad_norm": 0.25338777899742126, + "learning_rate": 5.4145445271346986e-05, + "loss": 0.0159, + "step": 5000 + }, + { + "epoch": 16.64451827242525, + "grad_norm": 0.25719591975212097, + "learning_rate": 5.398064519110622e-05, + "loss": 0.0142, + "step": 5010 + }, + { + "epoch": 16.677740863787374, + "grad_norm": 0.30168062448501587, + "learning_rate": 5.3815801579167394e-05, + "loss": 0.0138, + "step": 5020 + }, + { + "epoch": 16.710963455149503, + "grad_norm": 0.17436672747135162, + "learning_rate": 5.365091623823382e-05, + "loss": 0.0154, + "step": 5030 + }, + { + "epoch": 16.74418604651163, + "grad_norm": 0.23474833369255066, + "learning_rate": 5.348599097146521e-05, + "loss": 0.0153, + "step": 5040 + }, + { + "epoch": 16.777408637873755, + "grad_norm": 0.16410623490810394, + "learning_rate": 5.3321027582457836e-05, + "loss": 0.0117, + "step": 5050 + }, + { + "epoch": 16.81063122923588, + "grad_norm": 0.2162051796913147, + "learning_rate": 5.315602787522491e-05, + "loss": 0.0146, + "step": 5060 + }, + { + "epoch": 16.843853820598007, + "grad_norm": 0.25784435868263245, + "learning_rate": 5.299099365417678e-05, + "loss": 0.0136, + "step": 5070 + }, + { + "epoch": 16.877076411960132, + "grad_norm": 0.22715456783771515, + "learning_rate": 5.2825926724101236e-05, + "loss": 0.0129, + "step": 5080 + }, + { + "epoch": 16.910299003322258, + "grad_norm": 0.2097283899784088, + "learning_rate": 5.26608288901438e-05, + "loss": 0.0142, + "step": 5090 + }, + { + "epoch": 16.943521594684384, + "grad_norm": 0.2484341710805893, + "learning_rate": 5.24957019577879e-05, + "loss": 0.0142, + "step": 5100 + }, + { + "epoch": 16.97674418604651, + "grad_norm": 0.3111594021320343, + "learning_rate": 5.2330547732835266e-05, + "loss": 0.0171, + "step": 5110 + }, + { + "epoch": 17.00996677740864, + "grad_norm": 0.31417858600616455, + "learning_rate": 5.2165368021385996e-05, + "loss": 0.0208, + "step": 5120 + }, + { + "epoch": 17.043189368770765, + "grad_norm": 0.2889890968799591, + "learning_rate": 5.200016462981897e-05, + "loss": 0.0178, + "step": 5130 + }, + { + "epoch": 17.07641196013289, + "grad_norm": 0.2410278171300888, + "learning_rate": 5.1834939364772015e-05, + "loss": 0.0168, + "step": 5140 + }, + { + "epoch": 17.109634551495017, + "grad_norm": 0.22877810895442963, + "learning_rate": 5.166969403312214e-05, + "loss": 0.0143, + "step": 5150 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.3436548709869385, + "learning_rate": 5.1504430441965844e-05, + "loss": 0.0137, + "step": 5160 + }, + { + "epoch": 17.176079734219268, + "grad_norm": 0.1852232813835144, + "learning_rate": 5.133915039859923e-05, + "loss": 0.0144, + "step": 5170 + }, + { + "epoch": 17.209302325581394, + "grad_norm": 0.2806830108165741, + "learning_rate": 5.1173855710498444e-05, + "loss": 0.0125, + "step": 5180 + }, + { + "epoch": 17.242524916943523, + "grad_norm": 0.17646856606006622, + "learning_rate": 5.100854818529967e-05, + "loss": 0.012, + "step": 5190 + }, + { + "epoch": 17.27574750830565, + "grad_norm": 0.25370362401008606, + "learning_rate": 5.084322963077951e-05, + "loss": 0.015, + "step": 5200 + }, + { + "epoch": 17.308970099667775, + "grad_norm": 0.42946335673332214, + "learning_rate": 5.067790185483522e-05, + "loss": 0.0181, + "step": 5210 + }, + { + "epoch": 17.3421926910299, + "grad_norm": 0.2286258190870285, + "learning_rate": 5.0512566665464844e-05, + "loss": 0.0151, + "step": 5220 + }, + { + "epoch": 17.375415282392026, + "grad_norm": 0.25382083654403687, + "learning_rate": 5.034722587074755e-05, + "loss": 0.0142, + "step": 5230 + }, + { + "epoch": 17.408637873754152, + "grad_norm": 0.30657899379730225, + "learning_rate": 5.018188127882375e-05, + "loss": 0.0167, + "step": 5240 + }, + { + "epoch": 17.441860465116278, + "grad_norm": 0.27695420384407043, + "learning_rate": 5.0016534697875417e-05, + "loss": 0.0117, + "step": 5250 + }, + { + "epoch": 17.475083056478404, + "grad_norm": 0.18801657855510712, + "learning_rate": 4.9851187936106294e-05, + "loss": 0.0122, + "step": 5260 + }, + { + "epoch": 17.508305647840533, + "grad_norm": 0.23032335937023163, + "learning_rate": 4.968584280172206e-05, + "loss": 0.0156, + "step": 5270 + }, + { + "epoch": 17.54152823920266, + "grad_norm": 0.26843783259391785, + "learning_rate": 4.95205011029106e-05, + "loss": 0.0157, + "step": 5280 + }, + { + "epoch": 17.574750830564785, + "grad_norm": 0.3416936993598938, + "learning_rate": 4.935516464782227e-05, + "loss": 0.0155, + "step": 5290 + }, + { + "epoch": 17.60797342192691, + "grad_norm": 0.20719528198242188, + "learning_rate": 4.918983524455003e-05, + "loss": 0.013, + "step": 5300 + }, + { + "epoch": 17.641196013289036, + "grad_norm": 0.2207994908094406, + "learning_rate": 4.9024514701109766e-05, + "loss": 0.014, + "step": 5310 + }, + { + "epoch": 17.674418604651162, + "grad_norm": 0.25513842701911926, + "learning_rate": 4.885920482542043e-05, + "loss": 0.0133, + "step": 5320 + }, + { + "epoch": 17.707641196013288, + "grad_norm": 0.2589441239833832, + "learning_rate": 4.869390742528438e-05, + "loss": 0.0125, + "step": 5330 + }, + { + "epoch": 17.740863787375414, + "grad_norm": 0.28627756237983704, + "learning_rate": 4.852862430836744e-05, + "loss": 0.0137, + "step": 5340 + }, + { + "epoch": 17.774086378737543, + "grad_norm": 0.2636752128601074, + "learning_rate": 4.836335728217933e-05, + "loss": 0.0131, + "step": 5350 + }, + { + "epoch": 17.80730897009967, + "grad_norm": 0.19189167022705078, + "learning_rate": 4.819810815405379e-05, + "loss": 0.0133, + "step": 5360 + }, + { + "epoch": 17.840531561461795, + "grad_norm": 0.21339979767799377, + "learning_rate": 4.803287873112877e-05, + "loss": 0.0161, + "step": 5370 + }, + { + "epoch": 17.87375415282392, + "grad_norm": 0.27499207854270935, + "learning_rate": 4.786767082032681e-05, + "loss": 0.0144, + "step": 5380 + }, + { + "epoch": 17.906976744186046, + "grad_norm": 0.18827593326568604, + "learning_rate": 4.77024862283351e-05, + "loss": 0.0129, + "step": 5390 + }, + { + "epoch": 17.940199335548172, + "grad_norm": 0.15863770246505737, + "learning_rate": 4.753732676158593e-05, + "loss": 0.0129, + "step": 5400 + }, + { + "epoch": 17.973421926910298, + "grad_norm": 0.29626619815826416, + "learning_rate": 4.737219422623672e-05, + "loss": 0.0137, + "step": 5410 + }, + { + "epoch": 18.006644518272424, + "grad_norm": 0.19254808127880096, + "learning_rate": 4.720709042815044e-05, + "loss": 0.0169, + "step": 5420 + }, + { + "epoch": 18.039867109634553, + "grad_norm": 0.1836787462234497, + "learning_rate": 4.704201717287578e-05, + "loss": 0.0134, + "step": 5430 + }, + { + "epoch": 18.07308970099668, + "grad_norm": 0.20814216136932373, + "learning_rate": 4.6876976265627404e-05, + "loss": 0.0135, + "step": 5440 + }, + { + "epoch": 18.106312292358805, + "grad_norm": 0.17879468202590942, + "learning_rate": 4.671196951126626e-05, + "loss": 0.0138, + "step": 5450 + }, + { + "epoch": 18.13953488372093, + "grad_norm": 0.38961148262023926, + "learning_rate": 4.654699871427971e-05, + "loss": 0.0138, + "step": 5460 + }, + { + "epoch": 18.172757475083056, + "grad_norm": 0.23467040061950684, + "learning_rate": 4.6382065678762034e-05, + "loss": 0.0145, + "step": 5470 + }, + { + "epoch": 18.205980066445182, + "grad_norm": 0.23017248511314392, + "learning_rate": 4.6217172208394424e-05, + "loss": 0.0125, + "step": 5480 + }, + { + "epoch": 18.239202657807308, + "grad_norm": 0.25644469261169434, + "learning_rate": 4.605232010642549e-05, + "loss": 0.0132, + "step": 5490 + }, + { + "epoch": 18.272425249169434, + "grad_norm": 0.27478766441345215, + "learning_rate": 4.588751117565142e-05, + "loss": 0.0117, + "step": 5500 + }, + { + "epoch": 18.305647840531563, + "grad_norm": 0.25997769832611084, + "learning_rate": 4.5722747218396214e-05, + "loss": 0.0126, + "step": 5510 + }, + { + "epoch": 18.33887043189369, + "grad_norm": 0.2996649742126465, + "learning_rate": 4.5558030036492194e-05, + "loss": 0.0105, + "step": 5520 + }, + { + "epoch": 18.372093023255815, + "grad_norm": 0.13013173639774323, + "learning_rate": 4.539336143125999e-05, + "loss": 0.0136, + "step": 5530 + }, + { + "epoch": 18.40531561461794, + "grad_norm": 0.22051449120044708, + "learning_rate": 4.522874320348916e-05, + "loss": 0.0144, + "step": 5540 + }, + { + "epoch": 18.438538205980066, + "grad_norm": 0.24437175691127777, + "learning_rate": 4.506417715341821e-05, + "loss": 0.0199, + "step": 5550 + }, + { + "epoch": 18.471760797342192, + "grad_norm": 0.1537032574415207, + "learning_rate": 4.489966508071511e-05, + "loss": 0.0119, + "step": 5560 + }, + { + "epoch": 18.504983388704318, + "grad_norm": 0.1860104352235794, + "learning_rate": 4.4735208784457575e-05, + "loss": 0.011, + "step": 5570 + }, + { + "epoch": 18.538205980066444, + "grad_norm": 0.2010582834482193, + "learning_rate": 4.457081006311325e-05, + "loss": 0.0112, + "step": 5580 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.2242826372385025, + "learning_rate": 4.440647071452027e-05, + "loss": 0.012, + "step": 5590 + }, + { + "epoch": 18.6046511627907, + "grad_norm": 0.23328198492527008, + "learning_rate": 4.424219253586737e-05, + "loss": 0.0111, + "step": 5600 + }, + { + "epoch": 18.637873754152825, + "grad_norm": 0.19966065883636475, + "learning_rate": 4.407797732367443e-05, + "loss": 0.0145, + "step": 5610 + }, + { + "epoch": 18.67109634551495, + "grad_norm": 0.2838165760040283, + "learning_rate": 4.391382687377268e-05, + "loss": 0.0135, + "step": 5620 + }, + { + "epoch": 18.704318936877076, + "grad_norm": 0.2065529078245163, + "learning_rate": 4.374974298128512e-05, + "loss": 0.0114, + "step": 5630 + }, + { + "epoch": 18.737541528239202, + "grad_norm": 0.282686710357666, + "learning_rate": 4.358572744060699e-05, + "loss": 0.0129, + "step": 5640 + }, + { + "epoch": 18.770764119601328, + "grad_norm": 0.23530226945877075, + "learning_rate": 4.342178204538588e-05, + "loss": 0.0115, + "step": 5650 + }, + { + "epoch": 18.803986710963454, + "grad_norm": 0.19305096566677094, + "learning_rate": 4.325790858850241e-05, + "loss": 0.015, + "step": 5660 + }, + { + "epoch": 18.837209302325583, + "grad_norm": 0.2113550901412964, + "learning_rate": 4.309410886205043e-05, + "loss": 0.0096, + "step": 5670 + }, + { + "epoch": 18.87043189368771, + "grad_norm": 0.23037958145141602, + "learning_rate": 4.293038465731752e-05, + "loss": 0.0146, + "step": 5680 + }, + { + "epoch": 18.903654485049834, + "grad_norm": 0.18224431574344635, + "learning_rate": 4.276673776476533e-05, + "loss": 0.0121, + "step": 5690 + }, + { + "epoch": 18.93687707641196, + "grad_norm": 0.2601468563079834, + "learning_rate": 4.260316997401007e-05, + "loss": 0.013, + "step": 5700 + }, + { + "epoch": 18.970099667774086, + "grad_norm": 0.3363877832889557, + "learning_rate": 4.243968307380293e-05, + "loss": 0.015, + "step": 5710 + }, + { + "epoch": 19.003322259136212, + "grad_norm": 0.19996199011802673, + "learning_rate": 4.22762788520104e-05, + "loss": 0.0092, + "step": 5720 + }, + { + "epoch": 19.036544850498338, + "grad_norm": 0.20175497233867645, + "learning_rate": 4.211295909559491e-05, + "loss": 0.0158, + "step": 5730 + }, + { + "epoch": 19.069767441860463, + "grad_norm": 0.167605921626091, + "learning_rate": 4.194972559059511e-05, + "loss": 0.0124, + "step": 5740 + }, + { + "epoch": 19.102990033222593, + "grad_norm": 0.18364398181438446, + "learning_rate": 4.178658012210651e-05, + "loss": 0.0138, + "step": 5750 + }, + { + "epoch": 19.13621262458472, + "grad_norm": 0.29462730884552, + "learning_rate": 4.162352447426177e-05, + "loss": 0.0125, + "step": 5760 + }, + { + "epoch": 19.169435215946844, + "grad_norm": 0.2075863629579544, + "learning_rate": 4.146056043021135e-05, + "loss": 0.0116, + "step": 5770 + }, + { + "epoch": 19.20265780730897, + "grad_norm": 0.2390693575143814, + "learning_rate": 4.1297689772103944e-05, + "loss": 0.0137, + "step": 5780 + }, + { + "epoch": 19.235880398671096, + "grad_norm": 0.2695210576057434, + "learning_rate": 4.113491428106694e-05, + "loss": 0.0122, + "step": 5790 + }, + { + "epoch": 19.269102990033222, + "grad_norm": 0.23042955994606018, + "learning_rate": 4.0972235737187055e-05, + "loss": 0.014, + "step": 5800 + }, + { + "epoch": 19.302325581395348, + "grad_norm": 0.24274908006191254, + "learning_rate": 4.080965591949076e-05, + "loss": 0.0119, + "step": 5810 + }, + { + "epoch": 19.335548172757473, + "grad_norm": 0.256891131401062, + "learning_rate": 4.0647176605924924e-05, + "loss": 0.0096, + "step": 5820 + }, + { + "epoch": 19.368770764119603, + "grad_norm": 0.23432119190692902, + "learning_rate": 4.0484799573337255e-05, + "loss": 0.0121, + "step": 5830 + }, + { + "epoch": 19.40199335548173, + "grad_norm": 0.23672959208488464, + "learning_rate": 4.032252659745699e-05, + "loss": 0.0132, + "step": 5840 + }, + { + "epoch": 19.435215946843854, + "grad_norm": 0.18530480563640594, + "learning_rate": 4.016035945287539e-05, + "loss": 0.0126, + "step": 5850 + }, + { + "epoch": 19.46843853820598, + "grad_norm": 0.23715436458587646, + "learning_rate": 3.999829991302635e-05, + "loss": 0.013, + "step": 5860 + }, + { + "epoch": 19.501661129568106, + "grad_norm": 0.17024758458137512, + "learning_rate": 3.983634975016707e-05, + "loss": 0.0182, + "step": 5870 + }, + { + "epoch": 19.53488372093023, + "grad_norm": 0.23801341652870178, + "learning_rate": 3.967451073535854e-05, + "loss": 0.011, + "step": 5880 + }, + { + "epoch": 19.568106312292358, + "grad_norm": 0.20012985169887543, + "learning_rate": 3.951278463844633e-05, + "loss": 0.0117, + "step": 5890 + }, + { + "epoch": 19.601328903654483, + "grad_norm": 0.18276239931583405, + "learning_rate": 3.935117322804111e-05, + "loss": 0.0169, + "step": 5900 + }, + { + "epoch": 19.634551495016613, + "grad_norm": 0.32923269271850586, + "learning_rate": 3.918967827149938e-05, + "loss": 0.011, + "step": 5910 + }, + { + "epoch": 19.66777408637874, + "grad_norm": 0.12837934494018555, + "learning_rate": 3.9028301534904094e-05, + "loss": 0.0136, + "step": 5920 + }, + { + "epoch": 19.700996677740864, + "grad_norm": 0.14650031924247742, + "learning_rate": 3.88670447830454e-05, + "loss": 0.0099, + "step": 5930 + }, + { + "epoch": 19.73421926910299, + "grad_norm": 0.21855078637599945, + "learning_rate": 3.870590977940132e-05, + "loss": 0.0107, + "step": 5940 + }, + { + "epoch": 19.767441860465116, + "grad_norm": 0.30323830246925354, + "learning_rate": 3.8544898286118404e-05, + "loss": 0.0111, + "step": 5950 + }, + { + "epoch": 19.80066445182724, + "grad_norm": 0.23002244532108307, + "learning_rate": 3.838401206399257e-05, + "loss": 0.01, + "step": 5960 + }, + { + "epoch": 19.833887043189367, + "grad_norm": 0.1509399712085724, + "learning_rate": 3.822325287244975e-05, + "loss": 0.0094, + "step": 5970 + }, + { + "epoch": 19.867109634551497, + "grad_norm": 0.22663408517837524, + "learning_rate": 3.8062622469526725e-05, + "loss": 0.0103, + "step": 5980 + }, + { + "epoch": 19.900332225913623, + "grad_norm": 0.16443394124507904, + "learning_rate": 3.790212261185183e-05, + "loss": 0.0091, + "step": 5990 + }, + { + "epoch": 19.93355481727575, + "grad_norm": 0.226366326212883, + "learning_rate": 3.7741755054625794e-05, + "loss": 0.0119, + "step": 6000 + }, + { + "epoch": 19.966777408637874, + "grad_norm": 0.15423819422721863, + "learning_rate": 3.758152155160255e-05, + "loss": 0.0117, + "step": 6010 + }, + { + "epoch": 20.0, + "grad_norm": 0.5820074677467346, + "learning_rate": 3.742142385506999e-05, + "loss": 0.0103, + "step": 6020 + }, + { + "epoch": 20.033222591362126, + "grad_norm": 0.2385380119085312, + "learning_rate": 3.72614637158309e-05, + "loss": 0.0102, + "step": 6030 + }, + { + "epoch": 20.06644518272425, + "grad_norm": 0.25446420907974243, + "learning_rate": 3.710164288318371e-05, + "loss": 0.0117, + "step": 6040 + }, + { + "epoch": 20.099667774086377, + "grad_norm": 0.3364415168762207, + "learning_rate": 3.694196310490345e-05, + "loss": 0.0099, + "step": 6050 + }, + { + "epoch": 20.132890365448507, + "grad_norm": 0.23453129827976227, + "learning_rate": 3.678242612722259e-05, + "loss": 0.012, + "step": 6060 + }, + { + "epoch": 20.166112956810633, + "grad_norm": 0.19058457016944885, + "learning_rate": 3.6623033694811953e-05, + "loss": 0.0083, + "step": 6070 + }, + { + "epoch": 20.19933554817276, + "grad_norm": 0.21197690069675446, + "learning_rate": 3.6463787550761665e-05, + "loss": 0.0114, + "step": 6080 + }, + { + "epoch": 20.232558139534884, + "grad_norm": 0.2696029543876648, + "learning_rate": 3.630468943656202e-05, + "loss": 0.0095, + "step": 6090 + }, + { + "epoch": 20.26578073089701, + "grad_norm": 0.2509767413139343, + "learning_rate": 3.6145741092084523e-05, + "loss": 0.0131, + "step": 6100 + }, + { + "epoch": 20.299003322259136, + "grad_norm": 0.14039862155914307, + "learning_rate": 3.598694425556278e-05, + "loss": 0.0125, + "step": 6110 + }, + { + "epoch": 20.33222591362126, + "grad_norm": 0.11676931381225586, + "learning_rate": 3.58283006635736e-05, + "loss": 0.008, + "step": 6120 + }, + { + "epoch": 20.365448504983387, + "grad_norm": 0.2353682965040207, + "learning_rate": 3.566981205101781e-05, + "loss": 0.0106, + "step": 6130 + }, + { + "epoch": 20.398671096345517, + "grad_norm": 0.22952032089233398, + "learning_rate": 3.5511480151101556e-05, + "loss": 0.0115, + "step": 6140 + }, + { + "epoch": 20.431893687707642, + "grad_norm": 0.25359752774238586, + "learning_rate": 3.5353306695317104e-05, + "loss": 0.0118, + "step": 6150 + }, + { + "epoch": 20.46511627906977, + "grad_norm": 0.23889994621276855, + "learning_rate": 3.519529341342402e-05, + "loss": 0.0085, + "step": 6160 + }, + { + "epoch": 20.498338870431894, + "grad_norm": 0.11435896158218384, + "learning_rate": 3.503744203343026e-05, + "loss": 0.0079, + "step": 6170 + }, + { + "epoch": 20.53156146179402, + "grad_norm": 0.2550327777862549, + "learning_rate": 3.487975428157318e-05, + "loss": 0.0141, + "step": 6180 + }, + { + "epoch": 20.564784053156146, + "grad_norm": 0.15230843424797058, + "learning_rate": 3.472223188230083e-05, + "loss": 0.0119, + "step": 6190 + }, + { + "epoch": 20.59800664451827, + "grad_norm": 0.1876257359981537, + "learning_rate": 3.4564876558252866e-05, + "loss": 0.012, + "step": 6200 + }, + { + "epoch": 20.631229235880397, + "grad_norm": 0.2064155787229538, + "learning_rate": 3.440769003024195e-05, + "loss": 0.0121, + "step": 6210 + }, + { + "epoch": 20.664451827242527, + "grad_norm": 0.12937350571155548, + "learning_rate": 3.425067401723477e-05, + "loss": 0.0101, + "step": 6220 + }, + { + "epoch": 20.697674418604652, + "grad_norm": 0.2537197172641754, + "learning_rate": 3.409383023633325e-05, + "loss": 0.0125, + "step": 6230 + }, + { + "epoch": 20.730897009966778, + "grad_norm": 0.17130358517169952, + "learning_rate": 3.3937160402755894e-05, + "loss": 0.0118, + "step": 6240 + }, + { + "epoch": 20.764119601328904, + "grad_norm": 0.32221508026123047, + "learning_rate": 3.378066622981885e-05, + "loss": 0.0107, + "step": 6250 + }, + { + "epoch": 20.79734219269103, + "grad_norm": 0.13424603641033173, + "learning_rate": 3.362434942891738e-05, + "loss": 0.0112, + "step": 6260 + }, + { + "epoch": 20.830564784053156, + "grad_norm": 0.2683691382408142, + "learning_rate": 3.346821170950693e-05, + "loss": 0.0142, + "step": 6270 + }, + { + "epoch": 20.86378737541528, + "grad_norm": 0.28922441601753235, + "learning_rate": 3.3312254779084585e-05, + "loss": 0.0125, + "step": 6280 + }, + { + "epoch": 20.897009966777407, + "grad_norm": 0.3019663095474243, + "learning_rate": 3.315648034317039e-05, + "loss": 0.013, + "step": 6290 + }, + { + "epoch": 20.930232558139537, + "grad_norm": 0.2417183667421341, + "learning_rate": 3.3000890105288564e-05, + "loss": 0.0115, + "step": 6300 + }, + { + "epoch": 20.963455149501662, + "grad_norm": 0.12260009348392487, + "learning_rate": 3.284548576694908e-05, + "loss": 0.0135, + "step": 6310 + }, + { + "epoch": 20.996677740863788, + "grad_norm": 0.1857173889875412, + "learning_rate": 3.2690269027628815e-05, + "loss": 0.0105, + "step": 6320 + }, + { + "epoch": 21.029900332225914, + "grad_norm": 0.1845966875553131, + "learning_rate": 3.253524158475324e-05, + "loss": 0.012, + "step": 6330 + }, + { + "epoch": 21.06312292358804, + "grad_norm": 0.24921785295009613, + "learning_rate": 3.238040513367757e-05, + "loss": 0.0086, + "step": 6340 + }, + { + "epoch": 21.096345514950166, + "grad_norm": 0.27889764308929443, + "learning_rate": 3.222576136766843e-05, + "loss": 0.0122, + "step": 6350 + }, + { + "epoch": 21.12956810631229, + "grad_norm": 0.22039970755577087, + "learning_rate": 3.2071311977885324e-05, + "loss": 0.0088, + "step": 6360 + }, + { + "epoch": 21.162790697674417, + "grad_norm": 0.21786560118198395, + "learning_rate": 3.191705865336197e-05, + "loss": 0.0101, + "step": 6370 + }, + { + "epoch": 21.196013289036546, + "grad_norm": 0.2396402209997177, + "learning_rate": 3.1763003080988075e-05, + "loss": 0.0117, + "step": 6380 + }, + { + "epoch": 21.229235880398672, + "grad_norm": 0.21523918211460114, + "learning_rate": 3.160914694549063e-05, + "loss": 0.0118, + "step": 6390 + }, + { + "epoch": 21.262458471760798, + "grad_norm": 0.2706628143787384, + "learning_rate": 3.145549192941573e-05, + "loss": 0.0099, + "step": 6400 + }, + { + "epoch": 21.295681063122924, + "grad_norm": 0.25570225715637207, + "learning_rate": 3.130203971310999e-05, + "loss": 0.0112, + "step": 6410 + }, + { + "epoch": 21.32890365448505, + "grad_norm": 0.2580873370170593, + "learning_rate": 3.114879197470225e-05, + "loss": 0.0132, + "step": 6420 + }, + { + "epoch": 21.362126245847175, + "grad_norm": 0.23050130903720856, + "learning_rate": 3.0995750390085285e-05, + "loss": 0.0116, + "step": 6430 + }, + { + "epoch": 21.3953488372093, + "grad_norm": 0.2522919178009033, + "learning_rate": 3.084291663289728e-05, + "loss": 0.0101, + "step": 6440 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.12361680716276169, + "learning_rate": 3.069029237450375e-05, + "loss": 0.0084, + "step": 6450 + }, + { + "epoch": 21.461794019933556, + "grad_norm": 0.14684048295021057, + "learning_rate": 3.053787928397911e-05, + "loss": 0.0091, + "step": 6460 + }, + { + "epoch": 21.495016611295682, + "grad_norm": 0.2606411576271057, + "learning_rate": 3.0385679028088526e-05, + "loss": 0.0079, + "step": 6470 + }, + { + "epoch": 21.528239202657808, + "grad_norm": 0.20318882167339325, + "learning_rate": 3.023369327126959e-05, + "loss": 0.008, + "step": 6480 + }, + { + "epoch": 21.561461794019934, + "grad_norm": 0.20854884386062622, + "learning_rate": 3.0081923675614198e-05, + "loss": 0.0112, + "step": 6490 + }, + { + "epoch": 21.59468438538206, + "grad_norm": 0.19162316620349884, + "learning_rate": 2.993037190085034e-05, + "loss": 0.008, + "step": 6500 + }, + { + "epoch": 21.627906976744185, + "grad_norm": 0.20191577076911926, + "learning_rate": 2.977903960432392e-05, + "loss": 0.0088, + "step": 6510 + }, + { + "epoch": 21.66112956810631, + "grad_norm": 0.289853036403656, + "learning_rate": 2.9627928440980722e-05, + "loss": 0.0087, + "step": 6520 + }, + { + "epoch": 21.694352159468437, + "grad_norm": 0.16470327973365784, + "learning_rate": 2.9477040063348183e-05, + "loss": 0.011, + "step": 6530 + }, + { + "epoch": 21.727574750830566, + "grad_norm": 0.2899600565433502, + "learning_rate": 2.9326376121517456e-05, + "loss": 0.0128, + "step": 6540 + }, + { + "epoch": 21.760797342192692, + "grad_norm": 0.2802336812019348, + "learning_rate": 2.9175938263125236e-05, + "loss": 0.0138, + "step": 6550 + }, + { + "epoch": 21.794019933554818, + "grad_norm": 0.22840915620326996, + "learning_rate": 2.9025728133335873e-05, + "loss": 0.0117, + "step": 6560 + }, + { + "epoch": 21.827242524916944, + "grad_norm": 0.289288729429245, + "learning_rate": 2.8875747374823288e-05, + "loss": 0.0101, + "step": 6570 + }, + { + "epoch": 21.86046511627907, + "grad_norm": 0.19778776168823242, + "learning_rate": 2.872599762775298e-05, + "loss": 0.0102, + "step": 6580 + }, + { + "epoch": 21.893687707641195, + "grad_norm": 0.1731463372707367, + "learning_rate": 2.857648052976425e-05, + "loss": 0.0109, + "step": 6590 + }, + { + "epoch": 21.92691029900332, + "grad_norm": 0.14055639505386353, + "learning_rate": 2.8427197715952047e-05, + "loss": 0.0158, + "step": 6600 + }, + { + "epoch": 21.960132890365447, + "grad_norm": 0.22689878940582275, + "learning_rate": 2.8278150818849393e-05, + "loss": 0.0097, + "step": 6610 + }, + { + "epoch": 21.993355481727576, + "grad_norm": 0.17787189781665802, + "learning_rate": 2.812934146840922e-05, + "loss": 0.0111, + "step": 6620 + }, + { + "epoch": 22.026578073089702, + "grad_norm": 0.2983497977256775, + "learning_rate": 2.7980771291986764e-05, + "loss": 0.0128, + "step": 6630 + }, + { + "epoch": 22.059800664451828, + "grad_norm": 0.28037431836128235, + "learning_rate": 2.783244191432167e-05, + "loss": 0.0129, + "step": 6640 + }, + { + "epoch": 22.093023255813954, + "grad_norm": 0.17736779153347015, + "learning_rate": 2.768435495752022e-05, + "loss": 0.0143, + "step": 6650 + }, + { + "epoch": 22.12624584717608, + "grad_norm": 0.18235179781913757, + "learning_rate": 2.753651204103771e-05, + "loss": 0.0107, + "step": 6660 + }, + { + "epoch": 22.159468438538205, + "grad_norm": 0.21416562795639038, + "learning_rate": 2.7388914781660523e-05, + "loss": 0.0091, + "step": 6670 + }, + { + "epoch": 22.19269102990033, + "grad_norm": 0.22016173601150513, + "learning_rate": 2.7241564793488693e-05, + "loss": 0.0111, + "step": 6680 + }, + { + "epoch": 22.225913621262457, + "grad_norm": 0.20562654733657837, + "learning_rate": 2.7094463687918037e-05, + "loss": 0.0093, + "step": 6690 + }, + { + "epoch": 22.259136212624586, + "grad_norm": 0.21838662028312683, + "learning_rate": 2.694761307362268e-05, + "loss": 0.0127, + "step": 6700 + }, + { + "epoch": 22.292358803986712, + "grad_norm": 0.20718765258789062, + "learning_rate": 2.6801014556537467e-05, + "loss": 0.0092, + "step": 6710 + }, + { + "epoch": 22.325581395348838, + "grad_norm": 0.20617736876010895, + "learning_rate": 2.6654669739840243e-05, + "loss": 0.01, + "step": 6720 + }, + { + "epoch": 22.358803986710964, + "grad_norm": 0.24590416252613068, + "learning_rate": 2.650858022393451e-05, + "loss": 0.0097, + "step": 6730 + }, + { + "epoch": 22.39202657807309, + "grad_norm": 0.21153171360492706, + "learning_rate": 2.6362747606431747e-05, + "loss": 0.0112, + "step": 6740 + }, + { + "epoch": 22.425249169435215, + "grad_norm": 0.1685451716184616, + "learning_rate": 2.6217173482134172e-05, + "loss": 0.0116, + "step": 6750 + }, + { + "epoch": 22.45847176079734, + "grad_norm": 0.14820748567581177, + "learning_rate": 2.6071859443017044e-05, + "loss": 0.0119, + "step": 6760 + }, + { + "epoch": 22.491694352159467, + "grad_norm": 0.15515606105327606, + "learning_rate": 2.5926807078211414e-05, + "loss": 0.0101, + "step": 6770 + }, + { + "epoch": 22.524916943521596, + "grad_norm": 0.1173735037446022, + "learning_rate": 2.5782017973986728e-05, + "loss": 0.0061, + "step": 6780 + }, + { + "epoch": 22.558139534883722, + "grad_norm": 0.19914469122886658, + "learning_rate": 2.5637493713733374e-05, + "loss": 0.0097, + "step": 6790 + }, + { + "epoch": 22.591362126245848, + "grad_norm": 0.21056263148784637, + "learning_rate": 2.549323587794559e-05, + "loss": 0.0082, + "step": 6800 + }, + { + "epoch": 22.624584717607974, + "grad_norm": 0.1811390370130539, + "learning_rate": 2.5349246044203895e-05, + "loss": 0.0088, + "step": 6810 + }, + { + "epoch": 22.6578073089701, + "grad_norm": 0.24768255650997162, + "learning_rate": 2.520552578715808e-05, + "loss": 0.0107, + "step": 6820 + }, + { + "epoch": 22.691029900332225, + "grad_norm": 0.17765392363071442, + "learning_rate": 2.506207667850981e-05, + "loss": 0.0094, + "step": 6830 + }, + { + "epoch": 22.72425249169435, + "grad_norm": 0.23039370775222778, + "learning_rate": 2.4918900286995555e-05, + "loss": 0.0081, + "step": 6840 + }, + { + "epoch": 22.757475083056477, + "grad_norm": 0.1583041399717331, + "learning_rate": 2.4775998178369458e-05, + "loss": 0.008, + "step": 6850 + }, + { + "epoch": 22.790697674418606, + "grad_norm": 0.140150249004364, + "learning_rate": 2.4633371915386017e-05, + "loss": 0.0081, + "step": 6860 + }, + { + "epoch": 22.823920265780732, + "grad_norm": 0.12436725944280624, + "learning_rate": 2.4491023057783235e-05, + "loss": 0.0078, + "step": 6870 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.21081511676311493, + "learning_rate": 2.4348953162265375e-05, + "loss": 0.0086, + "step": 6880 + }, + { + "epoch": 22.890365448504983, + "grad_norm": 0.12631870806217194, + "learning_rate": 2.420716378248607e-05, + "loss": 0.0121, + "step": 6890 + }, + { + "epoch": 22.92358803986711, + "grad_norm": 0.11978428810834885, + "learning_rate": 2.4065656469031266e-05, + "loss": 0.0109, + "step": 6900 + }, + { + "epoch": 22.956810631229235, + "grad_norm": 0.14292840659618378, + "learning_rate": 2.3924432769402268e-05, + "loss": 0.0108, + "step": 6910 + }, + { + "epoch": 22.99003322259136, + "grad_norm": 0.17778891324996948, + "learning_rate": 2.3783494227998844e-05, + "loss": 0.0079, + "step": 6920 + }, + { + "epoch": 23.023255813953487, + "grad_norm": 0.18700818717479706, + "learning_rate": 2.3642842386102264e-05, + "loss": 0.0157, + "step": 6930 + }, + { + "epoch": 23.056478405315616, + "grad_norm": 0.22154302895069122, + "learning_rate": 2.3502478781858567e-05, + "loss": 0.016, + "step": 6940 + }, + { + "epoch": 23.089700996677742, + "grad_norm": 0.14165480434894562, + "learning_rate": 2.3362404950261628e-05, + "loss": 0.0084, + "step": 6950 + }, + { + "epoch": 23.122923588039868, + "grad_norm": 0.1183454692363739, + "learning_rate": 2.3222622423136458e-05, + "loss": 0.0089, + "step": 6960 + }, + { + "epoch": 23.156146179401993, + "grad_norm": 0.3031390905380249, + "learning_rate": 2.3083132729122332e-05, + "loss": 0.0123, + "step": 6970 + }, + { + "epoch": 23.18936877076412, + "grad_norm": 0.1480850726366043, + "learning_rate": 2.294393739365621e-05, + "loss": 0.0091, + "step": 6980 + }, + { + "epoch": 23.222591362126245, + "grad_norm": 0.19382554292678833, + "learning_rate": 2.2805037938956e-05, + "loss": 0.0074, + "step": 6990 + }, + { + "epoch": 23.25581395348837, + "grad_norm": 0.15653474628925323, + "learning_rate": 2.266643588400386e-05, + "loss": 0.0117, + "step": 7000 + }, + { + "epoch": 23.289036544850497, + "grad_norm": 0.18898485600948334, + "learning_rate": 2.252813274452969e-05, + "loss": 0.009, + "step": 7010 + }, + { + "epoch": 23.322259136212626, + "grad_norm": 0.14634114503860474, + "learning_rate": 2.2390130032994427e-05, + "loss": 0.0081, + "step": 7020 + }, + { + "epoch": 23.35548172757475, + "grad_norm": 0.1799151599407196, + "learning_rate": 2.2252429258573633e-05, + "loss": 0.0091, + "step": 7030 + }, + { + "epoch": 23.388704318936878, + "grad_norm": 0.13711026310920715, + "learning_rate": 2.2115031927140904e-05, + "loss": 0.0074, + "step": 7040 + }, + { + "epoch": 23.421926910299003, + "grad_norm": 0.25118833780288696, + "learning_rate": 2.1977939541251463e-05, + "loss": 0.0085, + "step": 7050 + }, + { + "epoch": 23.45514950166113, + "grad_norm": 0.24253642559051514, + "learning_rate": 2.1841153600125684e-05, + "loss": 0.0097, + "step": 7060 + }, + { + "epoch": 23.488372093023255, + "grad_norm": 0.2377505898475647, + "learning_rate": 2.170467559963267e-05, + "loss": 0.014, + "step": 7070 + }, + { + "epoch": 23.52159468438538, + "grad_norm": 0.15439948439598083, + "learning_rate": 2.1568507032273982e-05, + "loss": 0.009, + "step": 7080 + }, + { + "epoch": 23.55481727574751, + "grad_norm": 0.16242630779743195, + "learning_rate": 2.1432649387167264e-05, + "loss": 0.0092, + "step": 7090 + }, + { + "epoch": 23.588039867109636, + "grad_norm": 0.20561891794204712, + "learning_rate": 2.1297104150029973e-05, + "loss": 0.0081, + "step": 7100 + }, + { + "epoch": 23.62126245847176, + "grad_norm": 0.15706400573253632, + "learning_rate": 2.116187280316307e-05, + "loss": 0.0077, + "step": 7110 + }, + { + "epoch": 23.654485049833887, + "grad_norm": 0.22200746834278107, + "learning_rate": 2.1026956825434908e-05, + "loss": 0.0075, + "step": 7120 + }, + { + "epoch": 23.687707641196013, + "grad_norm": 0.17513321340084076, + "learning_rate": 2.0892357692265017e-05, + "loss": 0.0086, + "step": 7130 + }, + { + "epoch": 23.72093023255814, + "grad_norm": 0.15748685598373413, + "learning_rate": 2.0758076875607947e-05, + "loss": 0.0079, + "step": 7140 + }, + { + "epoch": 23.754152823920265, + "grad_norm": 0.2524207532405853, + "learning_rate": 2.0624115843937207e-05, + "loss": 0.0085, + "step": 7150 + }, + { + "epoch": 23.78737541528239, + "grad_norm": 0.1676102578639984, + "learning_rate": 2.0490476062229157e-05, + "loss": 0.009, + "step": 7160 + }, + { + "epoch": 23.82059800664452, + "grad_norm": 0.18977515399456024, + "learning_rate": 2.035715899194704e-05, + "loss": 0.0101, + "step": 7170 + }, + { + "epoch": 23.853820598006646, + "grad_norm": 0.16387055814266205, + "learning_rate": 2.022416609102499e-05, + "loss": 0.0097, + "step": 7180 + }, + { + "epoch": 23.88704318936877, + "grad_norm": 0.18103492259979248, + "learning_rate": 2.009149881385205e-05, + "loss": 0.0085, + "step": 7190 + }, + { + "epoch": 23.920265780730897, + "grad_norm": 0.17272917926311493, + "learning_rate": 1.995915861125634e-05, + "loss": 0.0087, + "step": 7200 + }, + { + "epoch": 23.953488372093023, + "grad_norm": 0.19008910655975342, + "learning_rate": 1.9827146930489065e-05, + "loss": 0.0116, + "step": 7210 + }, + { + "epoch": 23.98671096345515, + "grad_norm": 0.17901214957237244, + "learning_rate": 1.9695465215208848e-05, + "loss": 0.0084, + "step": 7220 + }, + { + "epoch": 24.019933554817275, + "grad_norm": 0.17212511599063873, + "learning_rate": 1.9564114905465813e-05, + "loss": 0.0064, + "step": 7230 + }, + { + "epoch": 24.0531561461794, + "grad_norm": 0.3507051467895508, + "learning_rate": 1.9433097437685936e-05, + "loss": 0.0068, + "step": 7240 + }, + { + "epoch": 24.08637873754153, + "grad_norm": 0.18519094586372375, + "learning_rate": 1.930241424465521e-05, + "loss": 0.0101, + "step": 7250 + }, + { + "epoch": 24.119601328903656, + "grad_norm": 0.24379052221775055, + "learning_rate": 1.9172066755504115e-05, + "loss": 0.0082, + "step": 7260 + }, + { + "epoch": 24.15282392026578, + "grad_norm": 0.12175963073968887, + "learning_rate": 1.9042056395691914e-05, + "loss": 0.0059, + "step": 7270 + }, + { + "epoch": 24.186046511627907, + "grad_norm": 0.21823808550834656, + "learning_rate": 1.8912384586991066e-05, + "loss": 0.0076, + "step": 7280 + }, + { + "epoch": 24.219269102990033, + "grad_norm": 0.16986721754074097, + "learning_rate": 1.8783052747471717e-05, + "loss": 0.007, + "step": 7290 + }, + { + "epoch": 24.25249169435216, + "grad_norm": 0.19669653475284576, + "learning_rate": 1.865406229148611e-05, + "loss": 0.0081, + "step": 7300 + }, + { + "epoch": 24.285714285714285, + "grad_norm": 0.14709433913230896, + "learning_rate": 1.8525414629653233e-05, + "loss": 0.0068, + "step": 7310 + }, + { + "epoch": 24.31893687707641, + "grad_norm": 0.1506931632757187, + "learning_rate": 1.8397111168843255e-05, + "loss": 0.0108, + "step": 7320 + }, + { + "epoch": 24.35215946843854, + "grad_norm": 0.14726154506206512, + "learning_rate": 1.8269153312162323e-05, + "loss": 0.0089, + "step": 7330 + }, + { + "epoch": 24.385382059800666, + "grad_norm": 0.21428826451301575, + "learning_rate": 1.8141542458937054e-05, + "loss": 0.0088, + "step": 7340 + }, + { + "epoch": 24.41860465116279, + "grad_norm": 0.13095128536224365, + "learning_rate": 1.8014280004699268e-05, + "loss": 0.0099, + "step": 7350 + }, + { + "epoch": 24.451827242524917, + "grad_norm": 0.17902006208896637, + "learning_rate": 1.788736734117078e-05, + "loss": 0.0103, + "step": 7360 + }, + { + "epoch": 24.485049833887043, + "grad_norm": 0.1946352869272232, + "learning_rate": 1.7760805856248152e-05, + "loss": 0.0072, + "step": 7370 + }, + { + "epoch": 24.51827242524917, + "grad_norm": 0.1819358617067337, + "learning_rate": 1.7634596933987518e-05, + "loss": 0.0072, + "step": 7380 + }, + { + "epoch": 24.551495016611295, + "grad_norm": 0.156645268201828, + "learning_rate": 1.7508741954589404e-05, + "loss": 0.0074, + "step": 7390 + }, + { + "epoch": 24.58471760797342, + "grad_norm": 0.16451464593410492, + "learning_rate": 1.7383242294383717e-05, + "loss": 0.0081, + "step": 7400 + }, + { + "epoch": 24.61794019933555, + "grad_norm": 0.12670625746250153, + "learning_rate": 1.7258099325814632e-05, + "loss": 0.0097, + "step": 7410 + }, + { + "epoch": 24.651162790697676, + "grad_norm": 0.14196054637432098, + "learning_rate": 1.7133314417425594e-05, + "loss": 0.0079, + "step": 7420 + }, + { + "epoch": 24.6843853820598, + "grad_norm": 0.11734014004468918, + "learning_rate": 1.7008888933844408e-05, + "loss": 0.0081, + "step": 7430 + }, + { + "epoch": 24.717607973421927, + "grad_norm": 0.11686446517705917, + "learning_rate": 1.6884824235768172e-05, + "loss": 0.0098, + "step": 7440 + }, + { + "epoch": 24.750830564784053, + "grad_norm": 0.21845193207263947, + "learning_rate": 1.6761121679948592e-05, + "loss": 0.0068, + "step": 7450 + }, + { + "epoch": 24.78405315614618, + "grad_norm": 0.19278839230537415, + "learning_rate": 1.663778261917695e-05, + "loss": 0.009, + "step": 7460 + }, + { + "epoch": 24.817275747508305, + "grad_norm": 0.13726863265037537, + "learning_rate": 1.651480840226952e-05, + "loss": 0.009, + "step": 7470 + }, + { + "epoch": 24.85049833887043, + "grad_norm": 0.14601445198059082, + "learning_rate": 1.639220037405258e-05, + "loss": 0.0085, + "step": 7480 + }, + { + "epoch": 24.88372093023256, + "grad_norm": 0.13516046106815338, + "learning_rate": 1.6269959875347906e-05, + "loss": 0.0106, + "step": 7490 + }, + { + "epoch": 24.916943521594686, + "grad_norm": 0.12706805765628815, + "learning_rate": 1.614808824295802e-05, + "loss": 0.0094, + "step": 7500 + }, + { + "epoch": 24.95016611295681, + "grad_norm": 0.15375488996505737, + "learning_rate": 1.602658680965152e-05, + "loss": 0.0078, + "step": 7510 + }, + { + "epoch": 24.983388704318937, + "grad_norm": 0.10782153159379959, + "learning_rate": 1.5905456904148686e-05, + "loss": 0.0067, + "step": 7520 + }, + { + "epoch": 25.016611295681063, + "grad_norm": 0.19712236523628235, + "learning_rate": 1.57846998511067e-05, + "loss": 0.0081, + "step": 7530 + }, + { + "epoch": 25.04983388704319, + "grad_norm": 0.1177951842546463, + "learning_rate": 1.566431697110538e-05, + "loss": 0.0089, + "step": 7540 + }, + { + "epoch": 25.083056478405314, + "grad_norm": 0.18782255053520203, + "learning_rate": 1.554430958063259e-05, + "loss": 0.0087, + "step": 7550 + }, + { + "epoch": 25.11627906976744, + "grad_norm": 0.1974753737449646, + "learning_rate": 1.5424678992069912e-05, + "loss": 0.0085, + "step": 7560 + }, + { + "epoch": 25.14950166112957, + "grad_norm": 0.27396559715270996, + "learning_rate": 1.5305426513678362e-05, + "loss": 0.0066, + "step": 7570 + }, + { + "epoch": 25.182724252491695, + "grad_norm": 0.15109151601791382, + "learning_rate": 1.518655344958388e-05, + "loss": 0.0075, + "step": 7580 + }, + { + "epoch": 25.21594684385382, + "grad_norm": 0.1766371876001358, + "learning_rate": 1.5068061099763275e-05, + "loss": 0.0059, + "step": 7590 + }, + { + "epoch": 25.249169435215947, + "grad_norm": 0.16913636028766632, + "learning_rate": 1.494995076002988e-05, + "loss": 0.0121, + "step": 7600 + }, + { + "epoch": 25.282392026578073, + "grad_norm": 0.15169569849967957, + "learning_rate": 1.4832223722019456e-05, + "loss": 0.0091, + "step": 7610 + }, + { + "epoch": 25.3156146179402, + "grad_norm": 0.1753765344619751, + "learning_rate": 1.4714881273176035e-05, + "loss": 0.0081, + "step": 7620 + }, + { + "epoch": 25.348837209302324, + "grad_norm": 0.16382759809494019, + "learning_rate": 1.4597924696737835e-05, + "loss": 0.0072, + "step": 7630 + }, + { + "epoch": 25.38205980066445, + "grad_norm": 0.23668570816516876, + "learning_rate": 1.4481355271723252e-05, + "loss": 0.0078, + "step": 7640 + }, + { + "epoch": 25.41528239202658, + "grad_norm": 0.153398334980011, + "learning_rate": 1.4365174272916809e-05, + "loss": 0.0071, + "step": 7650 + }, + { + "epoch": 25.448504983388705, + "grad_norm": 0.3028714656829834, + "learning_rate": 1.4249382970855319e-05, + "loss": 0.0073, + "step": 7660 + }, + { + "epoch": 25.48172757475083, + "grad_norm": 0.22805887460708618, + "learning_rate": 1.4133982631813903e-05, + "loss": 0.0077, + "step": 7670 + }, + { + "epoch": 25.514950166112957, + "grad_norm": 0.19053567945957184, + "learning_rate": 1.4018974517792194e-05, + "loss": 0.0087, + "step": 7680 + }, + { + "epoch": 25.548172757475083, + "grad_norm": 0.17864179611206055, + "learning_rate": 1.390435988650048e-05, + "loss": 0.0072, + "step": 7690 + }, + { + "epoch": 25.58139534883721, + "grad_norm": 0.13033527135849, + "learning_rate": 1.3790139991346006e-05, + "loss": 0.007, + "step": 7700 + }, + { + "epoch": 25.614617940199334, + "grad_norm": 0.12433844804763794, + "learning_rate": 1.367631608141926e-05, + "loss": 0.0081, + "step": 7710 + }, + { + "epoch": 25.64784053156146, + "grad_norm": 0.18164899945259094, + "learning_rate": 1.3562889401480278e-05, + "loss": 0.0065, + "step": 7720 + }, + { + "epoch": 25.68106312292359, + "grad_norm": 0.0895913615822792, + "learning_rate": 1.3449861191945074e-05, + "loss": 0.007, + "step": 7730 + }, + { + "epoch": 25.714285714285715, + "grad_norm": 0.21247518062591553, + "learning_rate": 1.3337232688872009e-05, + "loss": 0.0069, + "step": 7740 + }, + { + "epoch": 25.74750830564784, + "grad_norm": 0.20556800067424774, + "learning_rate": 1.3225005123948364e-05, + "loss": 0.0083, + "step": 7750 + }, + { + "epoch": 25.780730897009967, + "grad_norm": 0.15787535905838013, + "learning_rate": 1.311317972447681e-05, + "loss": 0.0089, + "step": 7760 + }, + { + "epoch": 25.813953488372093, + "grad_norm": 0.13835358619689941, + "learning_rate": 1.3001757713361996e-05, + "loss": 0.0073, + "step": 7770 + }, + { + "epoch": 25.84717607973422, + "grad_norm": 0.12107749283313751, + "learning_rate": 1.2890740309097204e-05, + "loss": 0.0073, + "step": 7780 + }, + { + "epoch": 25.880398671096344, + "grad_norm": 0.10240599513053894, + "learning_rate": 1.2780128725750944e-05, + "loss": 0.006, + "step": 7790 + }, + { + "epoch": 25.91362126245847, + "grad_norm": 0.14549055695533752, + "learning_rate": 1.266992417295379e-05, + "loss": 0.0067, + "step": 7800 + }, + { + "epoch": 25.9468438538206, + "grad_norm": 0.2025129497051239, + "learning_rate": 1.2560127855885073e-05, + "loss": 0.0063, + "step": 7810 + }, + { + "epoch": 25.980066445182725, + "grad_norm": 0.07502119243144989, + "learning_rate": 1.2450740975259745e-05, + "loss": 0.006, + "step": 7820 + }, + { + "epoch": 26.01328903654485, + "grad_norm": 0.14232605695724487, + "learning_rate": 1.234176472731517e-05, + "loss": 0.0054, + "step": 7830 + }, + { + "epoch": 26.046511627906977, + "grad_norm": 0.19360148906707764, + "learning_rate": 1.2233200303798158e-05, + "loss": 0.0103, + "step": 7840 + }, + { + "epoch": 26.079734219269103, + "grad_norm": 0.13156625628471375, + "learning_rate": 1.2125048891951846e-05, + "loss": 0.0081, + "step": 7850 + }, + { + "epoch": 26.11295681063123, + "grad_norm": 0.1898665726184845, + "learning_rate": 1.2017311674502745e-05, + "loss": 0.0111, + "step": 7860 + }, + { + "epoch": 26.146179401993354, + "grad_norm": 0.15508055686950684, + "learning_rate": 1.1909989829647822e-05, + "loss": 0.0071, + "step": 7870 + }, + { + "epoch": 26.17940199335548, + "grad_norm": 0.19671839475631714, + "learning_rate": 1.1803084531041553e-05, + "loss": 0.0054, + "step": 7880 + }, + { + "epoch": 26.21262458471761, + "grad_norm": 0.2435973882675171, + "learning_rate": 1.1696596947783162e-05, + "loss": 0.0081, + "step": 7890 + }, + { + "epoch": 26.245847176079735, + "grad_norm": 0.13469842076301575, + "learning_rate": 1.1590528244403803e-05, + "loss": 0.0084, + "step": 7900 + }, + { + "epoch": 26.27906976744186, + "grad_norm": 0.1078217402100563, + "learning_rate": 1.148487958085382e-05, + "loss": 0.008, + "step": 7910 + }, + { + "epoch": 26.312292358803987, + "grad_norm": 0.09761466830968857, + "learning_rate": 1.1379652112490086e-05, + "loss": 0.008, + "step": 7920 + }, + { + "epoch": 26.345514950166113, + "grad_norm": 0.22728003561496735, + "learning_rate": 1.1274846990063315e-05, + "loss": 0.0076, + "step": 7930 + }, + { + "epoch": 26.37873754152824, + "grad_norm": 0.1786966621875763, + "learning_rate": 1.117046535970554e-05, + "loss": 0.0084, + "step": 7940 + }, + { + "epoch": 26.411960132890364, + "grad_norm": 0.13671472668647766, + "learning_rate": 1.106650836291755e-05, + "loss": 0.0064, + "step": 7950 + }, + { + "epoch": 26.44518272425249, + "grad_norm": 0.1379915177822113, + "learning_rate": 1.0962977136556418e-05, + "loss": 0.0099, + "step": 7960 + }, + { + "epoch": 26.47840531561462, + "grad_norm": 0.2053348422050476, + "learning_rate": 1.0859872812823024e-05, + "loss": 0.0085, + "step": 7970 + }, + { + "epoch": 26.511627906976745, + "grad_norm": 0.15667298436164856, + "learning_rate": 1.0757196519249747e-05, + "loss": 0.0108, + "step": 7980 + }, + { + "epoch": 26.54485049833887, + "grad_norm": 0.13062669336795807, + "learning_rate": 1.0654949378688077e-05, + "loss": 0.0102, + "step": 7990 + }, + { + "epoch": 26.578073089700997, + "grad_norm": 0.17740964889526367, + "learning_rate": 1.0553132509296376e-05, + "loss": 0.0089, + "step": 8000 + }, + { + "epoch": 26.611295681063122, + "grad_norm": 0.15447011590003967, + "learning_rate": 1.0451747024527613e-05, + "loss": 0.0087, + "step": 8010 + }, + { + "epoch": 26.64451827242525, + "grad_norm": 0.14210566878318787, + "learning_rate": 1.0350794033117189e-05, + "loss": 0.0077, + "step": 8020 + }, + { + "epoch": 26.677740863787374, + "grad_norm": 0.1365930140018463, + "learning_rate": 1.0250274639070856e-05, + "loss": 0.0078, + "step": 8030 + }, + { + "epoch": 26.710963455149503, + "grad_norm": 0.1764291524887085, + "learning_rate": 1.0150189941652599e-05, + "loss": 0.0069, + "step": 8040 + }, + { + "epoch": 26.74418604651163, + "grad_norm": 0.1317356824874878, + "learning_rate": 1.0050541035372635e-05, + "loss": 0.0074, + "step": 8050 + }, + { + "epoch": 26.777408637873755, + "grad_norm": 0.1677209883928299, + "learning_rate": 9.951329009975458e-06, + "loss": 0.0072, + "step": 8060 + }, + { + "epoch": 26.81063122923588, + "grad_norm": 0.0969085842370987, + "learning_rate": 9.852554950427845e-06, + "loss": 0.0056, + "step": 8070 + }, + { + "epoch": 26.843853820598007, + "grad_norm": 0.11716752499341965, + "learning_rate": 9.754219936907105e-06, + "loss": 0.0054, + "step": 8080 + }, + { + "epoch": 26.877076411960132, + "grad_norm": 0.11322319507598877, + "learning_rate": 9.656325044789194e-06, + "loss": 0.0076, + "step": 8090 + }, + { + "epoch": 26.910299003322258, + "grad_norm": 0.08541762083768845, + "learning_rate": 9.55887134463697e-06, + "loss": 0.0059, + "step": 8100 + }, + { + "epoch": 26.943521594684384, + "grad_norm": 0.17191821336746216, + "learning_rate": 9.461859902188475e-06, + "loss": 0.0088, + "step": 8110 + }, + { + "epoch": 26.97674418604651, + "grad_norm": 0.1234305128455162, + "learning_rate": 9.365291778345303e-06, + "loss": 0.0046, + "step": 8120 + }, + { + "epoch": 27.00996677740864, + "grad_norm": 0.16608087718486786, + "learning_rate": 9.269168029160991e-06, + "loss": 0.0055, + "step": 8130 + }, + { + "epoch": 27.043189368770765, + "grad_norm": 0.22004766762256622, + "learning_rate": 9.173489705829447e-06, + "loss": 0.0077, + "step": 8140 + }, + { + "epoch": 27.07641196013289, + "grad_norm": 0.095779649913311, + "learning_rate": 9.078257854673516e-06, + "loss": 0.0066, + "step": 8150 + }, + { + "epoch": 27.109634551495017, + "grad_norm": 0.09602490812540054, + "learning_rate": 8.983473517133429e-06, + "loss": 0.0046, + "step": 8160 + }, + { + "epoch": 27.142857142857142, + "grad_norm": 0.11624102294445038, + "learning_rate": 8.889137729755537e-06, + "loss": 0.0084, + "step": 8170 + }, + { + "epoch": 27.176079734219268, + "grad_norm": 0.1401866227388382, + "learning_rate": 8.79525152418087e-06, + "loss": 0.0075, + "step": 8180 + }, + { + "epoch": 27.209302325581394, + "grad_norm": 0.10816841572523117, + "learning_rate": 8.701815927133961e-06, + "loss": 0.0054, + "step": 8190 + }, + { + "epoch": 27.242524916943523, + "grad_norm": 0.15892091393470764, + "learning_rate": 8.608831960411534e-06, + "loss": 0.007, + "step": 8200 + }, + { + "epoch": 27.27574750830565, + "grad_norm": 0.21621884405612946, + "learning_rate": 8.516300640871321e-06, + "loss": 0.0079, + "step": 8210 + }, + { + "epoch": 27.308970099667775, + "grad_norm": 0.14722590148448944, + "learning_rate": 8.424222980421038e-06, + "loss": 0.0067, + "step": 8220 + }, + { + "epoch": 27.3421926910299, + "grad_norm": 0.13478265702724457, + "learning_rate": 8.332599986007184e-06, + "loss": 0.0055, + "step": 8230 + }, + { + "epoch": 27.375415282392026, + "grad_norm": 0.22776460647583008, + "learning_rate": 8.241432659604203e-06, + "loss": 0.0086, + "step": 8240 + }, + { + "epoch": 27.408637873754152, + "grad_norm": 0.19308777153491974, + "learning_rate": 8.150721998203331e-06, + "loss": 0.0057, + "step": 8250 + }, + { + "epoch": 27.441860465116278, + "grad_norm": 0.14634957909584045, + "learning_rate": 8.06046899380184e-06, + "loss": 0.0072, + "step": 8260 + }, + { + "epoch": 27.475083056478404, + "grad_norm": 0.12301809340715408, + "learning_rate": 7.970674633392133e-06, + "loss": 0.0058, + "step": 8270 + }, + { + "epoch": 27.508305647840533, + "grad_norm": 0.09747922420501709, + "learning_rate": 7.881339898950924e-06, + "loss": 0.0068, + "step": 8280 + }, + { + "epoch": 27.54152823920266, + "grad_norm": 0.1105487197637558, + "learning_rate": 7.792465767428597e-06, + "loss": 0.0066, + "step": 8290 + }, + { + "epoch": 27.574750830564785, + "grad_norm": 0.2476223260164261, + "learning_rate": 7.704053210738376e-06, + "loss": 0.0104, + "step": 8300 + }, + { + "epoch": 27.60797342192691, + "grad_norm": 0.30715879797935486, + "learning_rate": 7.6161031957458494e-06, + "loss": 0.0063, + "step": 8310 + }, + { + "epoch": 27.641196013289036, + "grad_norm": 0.1410503387451172, + "learning_rate": 7.5286166842582605e-06, + "loss": 0.0071, + "step": 8320 + }, + { + "epoch": 27.674418604651162, + "grad_norm": 0.1309058517217636, + "learning_rate": 7.4415946330140814e-06, + "loss": 0.0059, + "step": 8330 + }, + { + "epoch": 27.707641196013288, + "grad_norm": 0.16256766021251678, + "learning_rate": 7.3550379936725644e-06, + "loss": 0.0089, + "step": 8340 + }, + { + "epoch": 27.740863787375414, + "grad_norm": 0.10909809172153473, + "learning_rate": 7.2689477128032035e-06, + "loss": 0.0052, + "step": 8350 + }, + { + "epoch": 27.774086378737543, + "grad_norm": 0.23795467615127563, + "learning_rate": 7.183324731875551e-06, + "loss": 0.0081, + "step": 8360 + }, + { + "epoch": 27.80730897009967, + "grad_norm": 0.1650657057762146, + "learning_rate": 7.098169987248782e-06, + "loss": 0.0067, + "step": 8370 + }, + { + "epoch": 27.840531561461795, + "grad_norm": 0.14811821281909943, + "learning_rate": 7.013484410161553e-06, + "loss": 0.006, + "step": 8380 + }, + { + "epoch": 27.87375415282392, + "grad_norm": 0.13227351009845734, + "learning_rate": 6.92926892672176e-06, + "loss": 0.0059, + "step": 8390 + }, + { + "epoch": 27.906976744186046, + "grad_norm": 0.1540694683790207, + "learning_rate": 6.845524457896446e-06, + "loss": 0.0058, + "step": 8400 + }, + { + "epoch": 27.940199335548172, + "grad_norm": 0.1506929099559784, + "learning_rate": 6.7622519195017165e-06, + "loss": 0.0064, + "step": 8410 + }, + { + "epoch": 27.973421926910298, + "grad_norm": 0.07793653756380081, + "learning_rate": 6.679452222192684e-06, + "loss": 0.0066, + "step": 8420 + }, + { + "epoch": 28.006644518272424, + "grad_norm": 0.2653924524784088, + "learning_rate": 6.597126271453579e-06, + "loss": 0.0035, + "step": 8430 + }, + { + "epoch": 28.039867109634553, + "grad_norm": 0.18384550511837006, + "learning_rate": 6.51527496758782e-06, + "loss": 0.0054, + "step": 8440 + }, + { + "epoch": 28.07308970099668, + "grad_norm": 0.11345284432172775, + "learning_rate": 6.433899205708155e-06, + "loss": 0.0048, + "step": 8450 + }, + { + "epoch": 28.106312292358805, + "grad_norm": 0.09172622859477997, + "learning_rate": 6.352999875726856e-06, + "loss": 0.005, + "step": 8460 + }, + { + "epoch": 28.13953488372093, + "grad_norm": 0.14451231062412262, + "learning_rate": 6.272577862346052e-06, + "loss": 0.005, + "step": 8470 + }, + { + "epoch": 28.172757475083056, + "grad_norm": 0.08029838651418686, + "learning_rate": 6.192634045047996e-06, + "loss": 0.0052, + "step": 8480 + }, + { + "epoch": 28.205980066445182, + "grad_norm": 0.12096463143825531, + "learning_rate": 6.113169298085458e-06, + "loss": 0.0063, + "step": 8490 + }, + { + "epoch": 28.239202657807308, + "grad_norm": 0.19323348999023438, + "learning_rate": 6.034184490472195e-06, + "loss": 0.0064, + "step": 8500 + }, + { + "epoch": 28.272425249169434, + "grad_norm": 0.07583433389663696, + "learning_rate": 5.955680485973386e-06, + "loss": 0.0116, + "step": 8510 + }, + { + "epoch": 28.305647840531563, + "grad_norm": 0.18973293900489807, + "learning_rate": 5.877658143096265e-06, + "loss": 0.0082, + "step": 8520 + }, + { + "epoch": 28.33887043189369, + "grad_norm": 0.1218564510345459, + "learning_rate": 5.800118315080661e-06, + "loss": 0.0062, + "step": 8530 + }, + { + "epoch": 28.372093023255815, + "grad_norm": 0.17067843675613403, + "learning_rate": 5.723061849889716e-06, + "loss": 0.0077, + "step": 8540 + }, + { + "epoch": 28.40531561461794, + "grad_norm": 0.18358756601810455, + "learning_rate": 5.646489590200604e-06, + "loss": 0.0061, + "step": 8550 + }, + { + "epoch": 28.438538205980066, + "grad_norm": 0.1261061728000641, + "learning_rate": 5.570402373395256e-06, + "loss": 0.0047, + "step": 8560 + }, + { + "epoch": 28.471760797342192, + "grad_norm": 0.08276877552270889, + "learning_rate": 5.494801031551305e-06, + "loss": 0.0055, + "step": 8570 + }, + { + "epoch": 28.504983388704318, + "grad_norm": 0.11832363903522491, + "learning_rate": 5.41968639143291e-06, + "loss": 0.0056, + "step": 8580 + }, + { + "epoch": 28.538205980066444, + "grad_norm": 0.17504636943340302, + "learning_rate": 5.345059274481751e-06, + "loss": 0.0054, + "step": 8590 + }, + { + "epoch": 28.571428571428573, + "grad_norm": 0.11633206903934479, + "learning_rate": 5.270920496808002e-06, + "loss": 0.0073, + "step": 8600 + }, + { + "epoch": 28.6046511627907, + "grad_norm": 0.16951923072338104, + "learning_rate": 5.1972708691814695e-06, + "loss": 0.0044, + "step": 8610 + }, + { + "epoch": 28.637873754152825, + "grad_norm": 0.06973368674516678, + "learning_rate": 5.124111197022674e-06, + "loss": 0.0068, + "step": 8620 + }, + { + "epoch": 28.67109634551495, + "grad_norm": 0.13720272481441498, + "learning_rate": 5.051442280394081e-06, + "loss": 0.0075, + "step": 8630 + }, + { + "epoch": 28.704318936877076, + "grad_norm": 0.12353954464197159, + "learning_rate": 4.979264913991322e-06, + "loss": 0.0073, + "step": 8640 + }, + { + "epoch": 28.737541528239202, + "grad_norm": 0.1034538745880127, + "learning_rate": 4.907579887134489e-06, + "loss": 0.0076, + "step": 8650 + }, + { + "epoch": 28.770764119601328, + "grad_norm": 0.1328953206539154, + "learning_rate": 4.836387983759572e-06, + "loss": 0.006, + "step": 8660 + }, + { + "epoch": 28.803986710963454, + "grad_norm": 0.13733744621276855, + "learning_rate": 4.765689982409816e-06, + "loss": 0.0056, + "step": 8670 + }, + { + "epoch": 28.837209302325583, + "grad_norm": 0.1435350924730301, + "learning_rate": 4.695486656227233e-06, + "loss": 0.0074, + "step": 8680 + }, + { + "epoch": 28.87043189368771, + "grad_norm": 0.10954142361879349, + "learning_rate": 4.625778772944156e-06, + "loss": 0.0056, + "step": 8690 + }, + { + "epoch": 28.903654485049834, + "grad_norm": 0.2844279408454895, + "learning_rate": 4.556567094874825e-06, + "loss": 0.006, + "step": 8700 + }, + { + "epoch": 28.93687707641196, + "grad_norm": 0.09259454160928726, + "learning_rate": 4.487852378907059e-06, + "loss": 0.0064, + "step": 8710 + }, + { + "epoch": 28.970099667774086, + "grad_norm": 0.11572569608688354, + "learning_rate": 4.419635376493986e-06, + "loss": 0.0049, + "step": 8720 + }, + { + "epoch": 29.003322259136212, + "grad_norm": 0.13336044549942017, + "learning_rate": 4.351916833645825e-06, + "loss": 0.0082, + "step": 8730 + }, + { + "epoch": 29.036544850498338, + "grad_norm": 0.09690898656845093, + "learning_rate": 4.284697490921691e-06, + "loss": 0.0039, + "step": 8740 + }, + { + "epoch": 29.069767441860463, + "grad_norm": 0.17928744852542877, + "learning_rate": 4.2179780834215585e-06, + "loss": 0.0065, + "step": 8750 + }, + { + "epoch": 29.102990033222593, + "grad_norm": 0.13231436908245087, + "learning_rate": 4.151759340778178e-06, + "loss": 0.0049, + "step": 8760 + }, + { + "epoch": 29.13621262458472, + "grad_norm": 0.19340196251869202, + "learning_rate": 4.086041987149109e-06, + "loss": 0.0079, + "step": 8770 + }, + { + "epoch": 29.169435215946844, + "grad_norm": 0.1559029370546341, + "learning_rate": 4.020826741208811e-06, + "loss": 0.0047, + "step": 8780 + }, + { + "epoch": 29.20265780730897, + "grad_norm": 0.12119326740503311, + "learning_rate": 3.956114316140746e-06, + "loss": 0.007, + "step": 8790 + }, + { + "epoch": 29.235880398671096, + "grad_norm": 0.14684447646141052, + "learning_rate": 3.891905419629643e-06, + "loss": 0.0055, + "step": 8800 + }, + { + "epoch": 29.269102990033222, + "grad_norm": 0.12230011075735092, + "learning_rate": 3.8282007538536946e-06, + "loss": 0.0067, + "step": 8810 + }, + { + "epoch": 29.302325581395348, + "grad_norm": 0.1537322849035263, + "learning_rate": 3.7650010154769265e-06, + "loss": 0.0073, + "step": 8820 + }, + { + "epoch": 29.335548172757473, + "grad_norm": 0.14978474378585815, + "learning_rate": 3.7023068956415608e-06, + "loss": 0.0097, + "step": 8830 + }, + { + "epoch": 29.368770764119603, + "grad_norm": 0.18187600374221802, + "learning_rate": 3.6401190799604303e-06, + "loss": 0.0044, + "step": 8840 + }, + { + "epoch": 29.40199335548173, + "grad_norm": 0.1751221865415573, + "learning_rate": 3.578438248509536e-06, + "loss": 0.0069, + "step": 8850 + }, + { + "epoch": 29.435215946843854, + "grad_norm": 0.12987150251865387, + "learning_rate": 3.5172650758205583e-06, + "loss": 0.0057, + "step": 8860 + }, + { + "epoch": 29.46843853820598, + "grad_norm": 0.07622663676738739, + "learning_rate": 3.45660023087353e-06, + "loss": 0.0046, + "step": 8870 + }, + { + "epoch": 29.501661129568106, + "grad_norm": 0.2524184286594391, + "learning_rate": 3.3964443770894528e-06, + "loss": 0.0088, + "step": 8880 + }, + { + "epoch": 29.53488372093023, + "grad_norm": 0.11506592482328415, + "learning_rate": 3.3367981723231245e-06, + "loss": 0.0066, + "step": 8890 + }, + { + "epoch": 29.568106312292358, + "grad_norm": 0.19349972903728485, + "learning_rate": 3.2776622688558746e-06, + "loss": 0.005, + "step": 8900 + }, + { + "epoch": 29.601328903654483, + "grad_norm": 0.16394150257110596, + "learning_rate": 3.2190373133884677e-06, + "loss": 0.006, + "step": 8910 + }, + { + "epoch": 29.634551495016613, + "grad_norm": 0.13305546343326569, + "learning_rate": 3.1609239470340446e-06, + "loss": 0.0051, + "step": 8920 + }, + { + "epoch": 29.66777408637874, + "grad_norm": 0.06419292092323303, + "learning_rate": 3.1033228053110373e-06, + "loss": 0.0082, + "step": 8930 + }, + { + "epoch": 29.700996677740864, + "grad_norm": 0.0762164518237114, + "learning_rate": 3.0462345181363314e-06, + "loss": 0.0051, + "step": 8940 + }, + { + "epoch": 29.73421926910299, + "grad_norm": 0.1778092384338379, + "learning_rate": 2.9896597098182654e-06, + "loss": 0.005, + "step": 8950 + }, + { + "epoch": 29.767441860465116, + "grad_norm": 0.21796007454395294, + "learning_rate": 2.933598999049891e-06, + "loss": 0.0053, + "step": 8960 + }, + { + "epoch": 29.80066445182724, + "grad_norm": 0.1496661901473999, + "learning_rate": 2.8780529989021697e-06, + "loss": 0.006, + "step": 8970 + }, + { + "epoch": 29.833887043189367, + "grad_norm": 0.1072162613272667, + "learning_rate": 2.823022316817242e-06, + "loss": 0.0069, + "step": 8980 + }, + { + "epoch": 29.867109634551497, + "grad_norm": 0.18262557685375214, + "learning_rate": 2.7685075546018456e-06, + "loss": 0.007, + "step": 8990 + }, + { + "epoch": 29.900332225913623, + "grad_norm": 0.16827644407749176, + "learning_rate": 2.7145093084206598e-06, + "loss": 0.0045, + "step": 9000 + }, + { + "epoch": 29.93355481727575, + "grad_norm": 0.1170128583908081, + "learning_rate": 2.661028168789892e-06, + "loss": 0.009, + "step": 9010 + }, + { + "epoch": 29.966777408637874, + "grad_norm": 0.18168750405311584, + "learning_rate": 2.6080647205706855e-06, + "loss": 0.0079, + "step": 9020 + }, + { + "epoch": 30.0, + "grad_norm": 0.13957791030406952, + "learning_rate": 2.555619542962834e-06, + "loss": 0.0063, + "step": 9030 + }, + { + "epoch": 30.033222591362126, + "grad_norm": 0.1216946691274643, + "learning_rate": 2.503693209498409e-06, + "loss": 0.0046, + "step": 9040 + }, + { + "epoch": 30.06644518272425, + "grad_norm": 0.10699830204248428, + "learning_rate": 2.452286288035449e-06, + "loss": 0.0047, + "step": 9050 + }, + { + "epoch": 30.099667774086377, + "grad_norm": 0.07205400615930557, + "learning_rate": 2.4013993407518363e-06, + "loss": 0.0041, + "step": 9060 + }, + { + "epoch": 30.132890365448507, + "grad_norm": 0.09559330344200134, + "learning_rate": 2.351032924139063e-06, + "loss": 0.0042, + "step": 9070 + }, + { + "epoch": 30.166112956810633, + "grad_norm": 0.10662047564983368, + "learning_rate": 2.30118758899619e-06, + "loss": 0.0064, + "step": 9080 + }, + { + "epoch": 30.19933554817276, + "grad_norm": 0.18922418355941772, + "learning_rate": 2.2518638804238157e-06, + "loss": 0.0062, + "step": 9090 + }, + { + "epoch": 30.232558139534884, + "grad_norm": 0.09248873591423035, + "learning_rate": 2.203062337818118e-06, + "loss": 0.0056, + "step": 9100 + }, + { + "epoch": 30.26578073089701, + "grad_norm": 0.07492813467979431, + "learning_rate": 2.1547834948649483e-06, + "loss": 0.0062, + "step": 9110 + }, + { + "epoch": 30.299003322259136, + "grad_norm": 0.04594607278704643, + "learning_rate": 2.1070278795340017e-06, + "loss": 0.0051, + "step": 9120 + }, + { + "epoch": 30.33222591362126, + "grad_norm": 0.1435636729001999, + "learning_rate": 2.059796014073029e-06, + "loss": 0.0065, + "step": 9130 + }, + { + "epoch": 30.365448504983387, + "grad_norm": 0.09786313772201538, + "learning_rate": 2.01308841500214e-06, + "loss": 0.006, + "step": 9140 + }, + { + "epoch": 30.398671096345517, + "grad_norm": 0.05220019444823265, + "learning_rate": 1.9669055931081704e-06, + "loss": 0.0063, + "step": 9150 + }, + { + "epoch": 30.431893687707642, + "grad_norm": 0.09716635942459106, + "learning_rate": 1.9212480534390507e-06, + "loss": 0.0037, + "step": 9160 + }, + { + "epoch": 30.46511627906977, + "grad_norm": 0.07489084452390671, + "learning_rate": 1.8761162952983246e-06, + "loss": 0.0055, + "step": 9170 + }, + { + "epoch": 30.498338870431894, + "grad_norm": 0.1921928972005844, + "learning_rate": 1.8315108122396618e-06, + "loss": 0.0084, + "step": 9180 + }, + { + "epoch": 30.53156146179402, + "grad_norm": 0.14873501658439636, + "learning_rate": 1.787432092061475e-06, + "loss": 0.0053, + "step": 9190 + }, + { + "epoch": 30.564784053156146, + "grad_norm": 0.18568743765354156, + "learning_rate": 1.743880616801602e-06, + "loss": 0.0074, + "step": 9200 + }, + { + "epoch": 30.59800664451827, + "grad_norm": 0.22179032862186432, + "learning_rate": 1.7008568627319865e-06, + "loss": 0.0058, + "step": 9210 + }, + { + "epoch": 30.631229235880397, + "grad_norm": 0.10541599243879318, + "learning_rate": 1.6583613003535226e-06, + "loss": 0.0046, + "step": 9220 + }, + { + "epoch": 30.664451827242527, + "grad_norm": 0.09513723850250244, + "learning_rate": 1.6163943943908522e-06, + "loss": 0.0067, + "step": 9230 + }, + { + "epoch": 30.697674418604652, + "grad_norm": 0.10265201330184937, + "learning_rate": 1.5749566037873476e-06, + "loss": 0.0047, + "step": 9240 + }, + { + "epoch": 30.730897009966778, + "grad_norm": 0.11902108788490295, + "learning_rate": 1.5340483817000428e-06, + "loss": 0.007, + "step": 9250 + }, + { + "epoch": 30.764119601328904, + "grad_norm": 0.12097522616386414, + "learning_rate": 1.4936701754947101e-06, + "loss": 0.0054, + "step": 9260 + }, + { + "epoch": 30.79734219269103, + "grad_norm": 0.09330645948648453, + "learning_rate": 1.4538224267409361e-06, + "loss": 0.0057, + "step": 9270 + }, + { + "epoch": 30.830564784053156, + "grad_norm": 0.1806746870279312, + "learning_rate": 1.414505571207314e-06, + "loss": 0.0086, + "step": 9280 + }, + { + "epoch": 30.86378737541528, + "grad_norm": 0.1468958705663681, + "learning_rate": 1.3757200388566816e-06, + "loss": 0.006, + "step": 9290 + }, + { + "epoch": 30.897009966777407, + "grad_norm": 0.032069865614175797, + "learning_rate": 1.3374662538414074e-06, + "loss": 0.0058, + "step": 9300 + }, + { + "epoch": 30.930232558139537, + "grad_norm": 0.1010846197605133, + "learning_rate": 1.2997446344987617e-06, + "loss": 0.0041, + "step": 9310 + }, + { + "epoch": 30.963455149501662, + "grad_norm": 0.06703367084264755, + "learning_rate": 1.262555593346315e-06, + "loss": 0.0058, + "step": 9320 + }, + { + "epoch": 30.996677740863788, + "grad_norm": 0.18721553683280945, + "learning_rate": 1.2258995370774685e-06, + "loss": 0.0074, + "step": 9330 + }, + { + "epoch": 31.029900332225914, + "grad_norm": 0.19383810460567474, + "learning_rate": 1.1897768665569798e-06, + "loss": 0.0068, + "step": 9340 + }, + { + "epoch": 31.06312292358804, + "grad_norm": 0.1317012906074524, + "learning_rate": 1.1541879768165954e-06, + "loss": 0.0059, + "step": 9350 + }, + { + "epoch": 31.096345514950166, + "grad_norm": 0.1485646516084671, + "learning_rate": 1.1191332570507085e-06, + "loss": 0.0074, + "step": 9360 + }, + { + "epoch": 31.12956810631229, + "grad_norm": 0.19343367218971252, + "learning_rate": 1.0846130906121132e-06, + "loss": 0.0065, + "step": 9370 + }, + { + "epoch": 31.162790697674417, + "grad_norm": 0.10918621718883514, + "learning_rate": 1.0506278550078131e-06, + "loss": 0.0048, + "step": 9380 + }, + { + "epoch": 31.196013289036546, + "grad_norm": 0.1376069337129593, + "learning_rate": 1.0171779218949185e-06, + "loss": 0.0075, + "step": 9390 + }, + { + "epoch": 31.229235880398672, + "grad_norm": 0.07469493895769119, + "learning_rate": 9.842636570765174e-07, + "loss": 0.0071, + "step": 9400 + }, + { + "epoch": 31.262458471760798, + "grad_norm": 0.0960230752825737, + "learning_rate": 9.518854204977612e-07, + "loss": 0.0065, + "step": 9410 + }, + { + "epoch": 31.295681063122924, + "grad_norm": 0.08492382615804672, + "learning_rate": 9.200435662418349e-07, + "loss": 0.0056, + "step": 9420 + }, + { + "epoch": 31.32890365448505, + "grad_norm": 0.11253948509693146, + "learning_rate": 8.887384425261658e-07, + "loss": 0.0053, + "step": 9430 + }, + { + "epoch": 31.362126245847175, + "grad_norm": 0.16154888272285461, + "learning_rate": 8.579703916985648e-07, + "loss": 0.0057, + "step": 9440 + }, + { + "epoch": 31.3953488372093, + "grad_norm": 0.14829444885253906, + "learning_rate": 8.277397502335194e-07, + "loss": 0.0059, + "step": 9450 + }, + { + "epoch": 31.428571428571427, + "grad_norm": 0.1551903337240219, + "learning_rate": 7.980468487284675e-07, + "loss": 0.0053, + "step": 9460 + }, + { + "epoch": 31.461794019933556, + "grad_norm": 0.1427416205406189, + "learning_rate": 7.688920119002297e-07, + "loss": 0.0081, + "step": 9470 + }, + { + "epoch": 31.495016611295682, + "grad_norm": 0.06173241510987282, + "learning_rate": 7.402755585814269e-07, + "loss": 0.0052, + "step": 9480 + }, + { + "epoch": 31.528239202657808, + "grad_norm": 0.20448872447013855, + "learning_rate": 7.121978017170073e-07, + "loss": 0.0065, + "step": 9490 + }, + { + "epoch": 31.561461794019934, + "grad_norm": 0.17006798088550568, + "learning_rate": 6.846590483608306e-07, + "loss": 0.0056, + "step": 9500 + }, + { + "epoch": 31.59468438538206, + "grad_norm": 0.1203543022274971, + "learning_rate": 6.576595996722834e-07, + "loss": 0.0055, + "step": 9510 + }, + { + "epoch": 31.627906976744185, + "grad_norm": 0.12413649260997772, + "learning_rate": 6.311997509130141e-07, + "loss": 0.0065, + "step": 9520 + }, + { + "epoch": 31.66112956810631, + "grad_norm": 0.2612221837043762, + "learning_rate": 6.052797914436803e-07, + "loss": 0.0093, + "step": 9530 + }, + { + "epoch": 31.694352159468437, + "grad_norm": 0.17194348573684692, + "learning_rate": 5.799000047208181e-07, + "loss": 0.0054, + "step": 9540 + }, + { + "epoch": 31.727574750830566, + "grad_norm": 0.11993417143821716, + "learning_rate": 5.550606682937054e-07, + "loss": 0.0068, + "step": 9550 + }, + { + "epoch": 31.760797342192692, + "grad_norm": 0.1255716234445572, + "learning_rate": 5.307620538013481e-07, + "loss": 0.0049, + "step": 9560 + }, + { + "epoch": 31.794019933554818, + "grad_norm": 0.1203114315867424, + "learning_rate": 5.070044269694874e-07, + "loss": 0.0086, + "step": 9570 + }, + { + "epoch": 31.827242524916944, + "grad_norm": 0.17764365673065186, + "learning_rate": 4.837880476077417e-07, + "loss": 0.0071, + "step": 9580 + }, + { + "epoch": 31.86046511627907, + "grad_norm": 0.1459934264421463, + "learning_rate": 4.6111316960670835e-07, + "loss": 0.0067, + "step": 9590 + }, + { + "epoch": 31.893687707641195, + "grad_norm": 0.24273799359798431, + "learning_rate": 4.389800409352218e-07, + "loss": 0.0065, + "step": 9600 + }, + { + "epoch": 31.92691029900332, + "grad_norm": 0.1205635517835617, + "learning_rate": 4.173889036376277e-07, + "loss": 0.0062, + "step": 9610 + }, + { + "epoch": 31.960132890365447, + "grad_norm": 0.08896120637655258, + "learning_rate": 3.963399938311463e-07, + "loss": 0.0048, + "step": 9620 + }, + { + "epoch": 31.993355481727576, + "grad_norm": 0.13939517736434937, + "learning_rate": 3.7583354170328545e-07, + "loss": 0.0043, + "step": 9630 + }, + { + "epoch": 32.0265780730897, + "grad_norm": 0.28227749466896057, + "learning_rate": 3.558697715093207e-07, + "loss": 0.0051, + "step": 9640 + }, + { + "epoch": 32.05980066445183, + "grad_norm": 0.11711577326059341, + "learning_rate": 3.3644890156983576e-07, + "loss": 0.0083, + "step": 9650 + }, + { + "epoch": 32.093023255813954, + "grad_norm": 0.188113734126091, + "learning_rate": 3.175711442683638e-07, + "loss": 0.0085, + "step": 9660 + }, + { + "epoch": 32.12624584717608, + "grad_norm": 0.15848992764949799, + "learning_rate": 2.9923670604902197e-07, + "loss": 0.0046, + "step": 9670 + }, + { + "epoch": 32.159468438538205, + "grad_norm": 0.09046372771263123, + "learning_rate": 2.814457874143028e-07, + "loss": 0.0055, + "step": 9680 + }, + { + "epoch": 32.19269102990033, + "grad_norm": 0.08672670274972916, + "learning_rate": 2.641985829228366e-07, + "loss": 0.0068, + "step": 9690 + }, + { + "epoch": 32.22591362126246, + "grad_norm": 0.22394372522830963, + "learning_rate": 2.474952811872877e-07, + "loss": 0.0063, + "step": 9700 + }, + { + "epoch": 32.25913621262458, + "grad_norm": 0.10601115971803665, + "learning_rate": 2.3133606487228397e-07, + "loss": 0.0065, + "step": 9710 + }, + { + "epoch": 32.29235880398671, + "grad_norm": 0.15347513556480408, + "learning_rate": 2.157211106924295e-07, + "loss": 0.0049, + "step": 9720 + }, + { + "epoch": 32.325581395348834, + "grad_norm": 0.2134745717048645, + "learning_rate": 2.006505894103672e-07, + "loss": 0.005, + "step": 9730 + }, + { + "epoch": 32.35880398671096, + "grad_norm": 0.15517526865005493, + "learning_rate": 1.8612466583489696e-07, + "loss": 0.0081, + "step": 9740 + }, + { + "epoch": 32.39202657807309, + "grad_norm": 0.07906592637300491, + "learning_rate": 1.7214349881918834e-07, + "loss": 0.0053, + "step": 9750 + }, + { + "epoch": 32.42524916943522, + "grad_norm": 0.17843244969844818, + "learning_rate": 1.5870724125904845e-07, + "loss": 0.007, + "step": 9760 + }, + { + "epoch": 32.458471760797345, + "grad_norm": 0.20996271073818207, + "learning_rate": 1.4581604009124006e-07, + "loss": 0.0057, + "step": 9770 + }, + { + "epoch": 32.49169435215947, + "grad_norm": 0.07877013832330704, + "learning_rate": 1.334700362918717e-07, + "loss": 0.0049, + "step": 9780 + }, + { + "epoch": 32.524916943521596, + "grad_norm": 0.18172283470630646, + "learning_rate": 1.2166936487486015e-07, + "loss": 0.0046, + "step": 9790 + }, + { + "epoch": 32.55813953488372, + "grad_norm": 0.082504041492939, + "learning_rate": 1.1041415489045914e-07, + "loss": 0.0049, + "step": 9800 + }, + { + "epoch": 32.59136212624585, + "grad_norm": 0.05255156010389328, + "learning_rate": 9.970452942384412e-08, + "loss": 0.0051, + "step": 9810 + }, + { + "epoch": 32.62458471760797, + "grad_norm": 0.07134050875902176, + "learning_rate": 8.954060559375754e-08, + "loss": 0.008, + "step": 9820 + }, + { + "epoch": 32.6578073089701, + "grad_norm": 0.09449620544910431, + "learning_rate": 7.99224945512489e-08, + "loss": 0.0044, + "step": 9830 + }, + { + "epoch": 32.691029900332225, + "grad_norm": 0.10606354475021362, + "learning_rate": 7.085030147843675e-08, + "loss": 0.0064, + "step": 9840 + }, + { + "epoch": 32.72425249169435, + "grad_norm": 0.10147126019001007, + "learning_rate": 6.232412558736523e-08, + "loss": 0.0037, + "step": 9850 + }, + { + "epoch": 32.75747508305648, + "grad_norm": 0.12870968878269196, + "learning_rate": 5.434406011893822e-08, + "loss": 0.0094, + "step": 9860 + }, + { + "epoch": 32.7906976744186, + "grad_norm": 0.09744128584861755, + "learning_rate": 4.6910192341864664e-08, + "loss": 0.0063, + "step": 9870 + }, + { + "epoch": 32.82392026578073, + "grad_norm": 0.13845191895961761, + "learning_rate": 4.0022603551737035e-08, + "loss": 0.0071, + "step": 9880 + }, + { + "epoch": 32.857142857142854, + "grad_norm": 0.08200283348560333, + "learning_rate": 3.3681369070120985e-08, + "loss": 0.0078, + "step": 9890 + }, + { + "epoch": 32.89036544850498, + "grad_norm": 0.08790437132120132, + "learning_rate": 2.7886558243744866e-08, + "loss": 0.004, + "step": 9900 + }, + { + "epoch": 32.92358803986711, + "grad_norm": 0.13174009323120117, + "learning_rate": 2.2638234443722596e-08, + "loss": 0.0061, + "step": 9910 + }, + { + "epoch": 32.95681063122924, + "grad_norm": 0.20475421845912933, + "learning_rate": 1.7936455064887504e-08, + "loss": 0.008, + "step": 9920 + }, + { + "epoch": 32.990033222591364, + "grad_norm": 0.12375841289758682, + "learning_rate": 1.378127152514841e-08, + "loss": 0.0069, + "step": 9930 + }, + { + "epoch": 33.02325581395349, + "grad_norm": 0.11901507526636124, + "learning_rate": 1.0172729264917857e-08, + "loss": 0.0056, + "step": 9940 + }, + { + "epoch": 33.056478405315616, + "grad_norm": 0.049505483359098434, + "learning_rate": 7.1108677466458215e-09, + "loss": 0.0094, + "step": 9950 + }, + { + "epoch": 33.08970099667774, + "grad_norm": 0.08091825991868973, + "learning_rate": 4.595720454353414e-09, + "loss": 0.0049, + "step": 9960 + }, + { + "epoch": 33.12292358803987, + "grad_norm": 0.12687945365905762, + "learning_rate": 2.627314893294264e-09, + "loss": 0.0081, + "step": 9970 + }, + { + "epoch": 33.15614617940199, + "grad_norm": 0.09400459378957748, + "learning_rate": 1.2056725896270048e-09, + "loss": 0.0069, + "step": 9980 + }, + { + "epoch": 33.18936877076412, + "grad_norm": 0.06770706921815872, + "learning_rate": 3.308090902098826e-10, + "loss": 0.0068, + "step": 9990 + }, + { + "epoch": 33.222591362126245, + "grad_norm": 0.18788298964500427, + "learning_rate": 2.7339624120159555e-12, + "loss": 0.0052, + "step": 10000 + } + ], + "logging_steps": 10, + "max_steps": 10000, + "num_input_tokens_seen": 0, + "num_train_epochs": 34, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +}