diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,28033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.732327713694174, + "eval_steps": 500, + "global_step": 4000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0011830819284235432, + "grad_norm": 3.1850123405456543, + "learning_rate": 4.99875e-06, + "loss": 0.3592, + "step": 1 + }, + { + "epoch": 0.0023661638568470865, + "grad_norm": 2.561338186264038, + "learning_rate": 4.997500000000001e-06, + "loss": 0.3357, + "step": 2 + }, + { + "epoch": 0.00354924578527063, + "grad_norm": 2.353553295135498, + "learning_rate": 4.996250000000001e-06, + "loss": 0.3063, + "step": 3 + }, + { + "epoch": 0.004732327713694173, + "grad_norm": 2.684662103652954, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.3014, + "step": 4 + }, + { + "epoch": 0.005915409642117717, + "grad_norm": 2.3714146614074707, + "learning_rate": 4.99375e-06, + "loss": 0.2829, + "step": 5 + }, + { + "epoch": 0.00709849157054126, + "grad_norm": 2.2721121311187744, + "learning_rate": 4.992500000000001e-06, + "loss": 0.279, + "step": 6 + }, + { + "epoch": 0.008281573498964804, + "grad_norm": 2.861663579940796, + "learning_rate": 4.99125e-06, + "loss": 0.3021, + "step": 7 + }, + { + "epoch": 0.009464655427388346, + "grad_norm": 2.373079538345337, + "learning_rate": 4.9900000000000005e-06, + "loss": 0.2663, + "step": 8 + }, + { + "epoch": 0.01064773735581189, + "grad_norm": 2.7103497982025146, + "learning_rate": 4.98875e-06, + "loss": 0.2822, + "step": 9 + }, + { + "epoch": 0.011830819284235433, + "grad_norm": 2.3081657886505127, + "learning_rate": 4.987500000000001e-06, + "loss": 0.2744, + "step": 10 + }, + { + "epoch": 0.013013901212658977, + "grad_norm": 2.174342393875122, + "learning_rate": 4.98625e-06, + "loss": 0.2654, + "step": 11 + }, + { + "epoch": 0.01419698314108252, + "grad_norm": 2.2082788944244385, + "learning_rate": 4.9850000000000006e-06, + "loss": 0.2634, + "step": 12 + }, + { + "epoch": 0.015380065069506064, + "grad_norm": 2.860771656036377, + "learning_rate": 4.98375e-06, + "loss": 0.2864, + "step": 13 + }, + { + "epoch": 0.016563146997929608, + "grad_norm": 2.8682806491851807, + "learning_rate": 4.982500000000001e-06, + "loss": 0.256, + "step": 14 + }, + { + "epoch": 0.01774622892635315, + "grad_norm": 2.2953498363494873, + "learning_rate": 4.98125e-06, + "loss": 0.2762, + "step": 15 + }, + { + "epoch": 0.01892931085477669, + "grad_norm": 2.03486967086792, + "learning_rate": 4.980000000000001e-06, + "loss": 0.2664, + "step": 16 + }, + { + "epoch": 0.020112392783200237, + "grad_norm": 2.0175719261169434, + "learning_rate": 4.97875e-06, + "loss": 0.2414, + "step": 17 + }, + { + "epoch": 0.02129547471162378, + "grad_norm": 2.520503282546997, + "learning_rate": 4.977500000000001e-06, + "loss": 0.2663, + "step": 18 + }, + { + "epoch": 0.022478556640047324, + "grad_norm": 2.578657865524292, + "learning_rate": 4.97625e-06, + "loss": 0.2729, + "step": 19 + }, + { + "epoch": 0.023661638568470866, + "grad_norm": 2.1403181552886963, + "learning_rate": 4.975000000000001e-06, + "loss": 0.2512, + "step": 20 + }, + { + "epoch": 0.024844720496894408, + "grad_norm": 2.019257068634033, + "learning_rate": 4.9737500000000004e-06, + "loss": 0.2221, + "step": 21 + }, + { + "epoch": 0.026027802425317954, + "grad_norm": 2.421431303024292, + "learning_rate": 4.9725e-06, + "loss": 0.2652, + "step": 22 + }, + { + "epoch": 0.027210884353741496, + "grad_norm": 1.9099886417388916, + "learning_rate": 4.97125e-06, + "loss": 0.2469, + "step": 23 + }, + { + "epoch": 0.02839396628216504, + "grad_norm": 2.1174657344818115, + "learning_rate": 4.970000000000001e-06, + "loss": 0.2446, + "step": 24 + }, + { + "epoch": 0.029577048210588583, + "grad_norm": 1.9696910381317139, + "learning_rate": 4.9687500000000005e-06, + "loss": 0.2277, + "step": 25 + }, + { + "epoch": 0.03076013013901213, + "grad_norm": 2.0108089447021484, + "learning_rate": 4.9675e-06, + "loss": 0.2459, + "step": 26 + }, + { + "epoch": 0.03194321206743567, + "grad_norm": 2.2769222259521484, + "learning_rate": 4.96625e-06, + "loss": 0.2515, + "step": 27 + }, + { + "epoch": 0.033126293995859216, + "grad_norm": 2.201805830001831, + "learning_rate": 4.965000000000001e-06, + "loss": 0.2567, + "step": 28 + }, + { + "epoch": 0.03430937592428276, + "grad_norm": 2.263960599899292, + "learning_rate": 4.9637500000000005e-06, + "loss": 0.2563, + "step": 29 + }, + { + "epoch": 0.0354924578527063, + "grad_norm": 1.9786373376846313, + "learning_rate": 4.9625e-06, + "loss": 0.2364, + "step": 30 + }, + { + "epoch": 0.03667553978112984, + "grad_norm": 2.3260087966918945, + "learning_rate": 4.96125e-06, + "loss": 0.2428, + "step": 31 + }, + { + "epoch": 0.03785862170955338, + "grad_norm": 2.4209721088409424, + "learning_rate": 4.960000000000001e-06, + "loss": 0.246, + "step": 32 + }, + { + "epoch": 0.03904170363797693, + "grad_norm": 2.0027618408203125, + "learning_rate": 4.958750000000001e-06, + "loss": 0.2551, + "step": 33 + }, + { + "epoch": 0.040224785566400474, + "grad_norm": 2.0058484077453613, + "learning_rate": 4.9575e-06, + "loss": 0.2177, + "step": 34 + }, + { + "epoch": 0.041407867494824016, + "grad_norm": 1.8714518547058105, + "learning_rate": 4.95625e-06, + "loss": 0.2294, + "step": 35 + }, + { + "epoch": 0.04259094942324756, + "grad_norm": 2.051802158355713, + "learning_rate": 4.955e-06, + "loss": 0.2643, + "step": 36 + }, + { + "epoch": 0.0437740313516711, + "grad_norm": 2.301116466522217, + "learning_rate": 4.953750000000001e-06, + "loss": 0.246, + "step": 37 + }, + { + "epoch": 0.04495711328009465, + "grad_norm": 2.205218553543091, + "learning_rate": 4.9525000000000004e-06, + "loss": 0.245, + "step": 38 + }, + { + "epoch": 0.04614019520851819, + "grad_norm": 2.1235976219177246, + "learning_rate": 4.95125e-06, + "loss": 0.2475, + "step": 39 + }, + { + "epoch": 0.04732327713694173, + "grad_norm": 2.3169736862182617, + "learning_rate": 4.95e-06, + "loss": 0.2669, + "step": 40 + }, + { + "epoch": 0.048506359065365275, + "grad_norm": 1.941275715827942, + "learning_rate": 4.948750000000001e-06, + "loss": 0.2335, + "step": 41 + }, + { + "epoch": 0.049689440993788817, + "grad_norm": 2.2764127254486084, + "learning_rate": 4.9475000000000005e-06, + "loss": 0.2472, + "step": 42 + }, + { + "epoch": 0.050872522922212365, + "grad_norm": 2.1738240718841553, + "learning_rate": 4.94625e-06, + "loss": 0.237, + "step": 43 + }, + { + "epoch": 0.05205560485063591, + "grad_norm": 2.2620325088500977, + "learning_rate": 4.945e-06, + "loss": 0.27, + "step": 44 + }, + { + "epoch": 0.05323868677905945, + "grad_norm": 2.079681158065796, + "learning_rate": 4.943750000000001e-06, + "loss": 0.2721, + "step": 45 + }, + { + "epoch": 0.05442176870748299, + "grad_norm": 2.0338194370269775, + "learning_rate": 4.9425000000000005e-06, + "loss": 0.2512, + "step": 46 + }, + { + "epoch": 0.05560485063590653, + "grad_norm": 2.0005125999450684, + "learning_rate": 4.94125e-06, + "loss": 0.2566, + "step": 47 + }, + { + "epoch": 0.05678793256433008, + "grad_norm": 2.0430355072021484, + "learning_rate": 4.94e-06, + "loss": 0.2498, + "step": 48 + }, + { + "epoch": 0.057971014492753624, + "grad_norm": 2.114612579345703, + "learning_rate": 4.938750000000001e-06, + "loss": 0.2396, + "step": 49 + }, + { + "epoch": 0.059154096421177166, + "grad_norm": 1.7274242639541626, + "learning_rate": 4.937500000000001e-06, + "loss": 0.2063, + "step": 50 + }, + { + "epoch": 0.06033717834960071, + "grad_norm": 4.613703727722168, + "learning_rate": 4.93625e-06, + "loss": 0.2515, + "step": 51 + }, + { + "epoch": 0.06152026027802426, + "grad_norm": 1.973289132118225, + "learning_rate": 4.935e-06, + "loss": 0.2322, + "step": 52 + }, + { + "epoch": 0.0627033422064478, + "grad_norm": 2.0583972930908203, + "learning_rate": 4.933750000000001e-06, + "loss": 0.2599, + "step": 53 + }, + { + "epoch": 0.06388642413487133, + "grad_norm": 2.534259557723999, + "learning_rate": 4.932500000000001e-06, + "loss": 0.239, + "step": 54 + }, + { + "epoch": 0.06506950606329488, + "grad_norm": 2.577638626098633, + "learning_rate": 4.9312500000000004e-06, + "loss": 0.2176, + "step": 55 + }, + { + "epoch": 0.06625258799171843, + "grad_norm": 2.2335028648376465, + "learning_rate": 4.93e-06, + "loss": 0.2492, + "step": 56 + }, + { + "epoch": 0.06743566992014197, + "grad_norm": 1.9347782135009766, + "learning_rate": 4.928750000000001e-06, + "loss": 0.2129, + "step": 57 + }, + { + "epoch": 0.06861875184856552, + "grad_norm": 2.321319818496704, + "learning_rate": 4.927500000000001e-06, + "loss": 0.2359, + "step": 58 + }, + { + "epoch": 0.06980183377698905, + "grad_norm": 2.1429150104522705, + "learning_rate": 4.9262500000000005e-06, + "loss": 0.2619, + "step": 59 + }, + { + "epoch": 0.0709849157054126, + "grad_norm": 2.657050609588623, + "learning_rate": 4.925e-06, + "loss": 0.2559, + "step": 60 + }, + { + "epoch": 0.07216799763383615, + "grad_norm": 2.193398952484131, + "learning_rate": 4.92375e-06, + "loss": 0.2461, + "step": 61 + }, + { + "epoch": 0.07335107956225968, + "grad_norm": 2.1088688373565674, + "learning_rate": 4.922500000000001e-06, + "loss": 0.2234, + "step": 62 + }, + { + "epoch": 0.07453416149068323, + "grad_norm": 2.0798721313476562, + "learning_rate": 4.9212500000000005e-06, + "loss": 0.2324, + "step": 63 + }, + { + "epoch": 0.07571724341910677, + "grad_norm": 1.990342140197754, + "learning_rate": 4.92e-06, + "loss": 0.2597, + "step": 64 + }, + { + "epoch": 0.07690032534753032, + "grad_norm": 2.122241258621216, + "learning_rate": 4.91875e-06, + "loss": 0.2345, + "step": 65 + }, + { + "epoch": 0.07808340727595386, + "grad_norm": 2.0237135887145996, + "learning_rate": 4.917500000000001e-06, + "loss": 0.2486, + "step": 66 + }, + { + "epoch": 0.0792664892043774, + "grad_norm": 2.1832430362701416, + "learning_rate": 4.9162500000000006e-06, + "loss": 0.2503, + "step": 67 + }, + { + "epoch": 0.08044957113280095, + "grad_norm": 2.299929618835449, + "learning_rate": 4.915e-06, + "loss": 0.2592, + "step": 68 + }, + { + "epoch": 0.08163265306122448, + "grad_norm": 2.149385929107666, + "learning_rate": 4.91375e-06, + "loss": 0.2344, + "step": 69 + }, + { + "epoch": 0.08281573498964803, + "grad_norm": 2.2965173721313477, + "learning_rate": 4.912500000000001e-06, + "loss": 0.2532, + "step": 70 + }, + { + "epoch": 0.08399881691807158, + "grad_norm": 2.102252960205078, + "learning_rate": 4.91125e-06, + "loss": 0.2544, + "step": 71 + }, + { + "epoch": 0.08518189884649512, + "grad_norm": 2.077648162841797, + "learning_rate": 4.9100000000000004e-06, + "loss": 0.2424, + "step": 72 + }, + { + "epoch": 0.08636498077491866, + "grad_norm": 1.7898882627487183, + "learning_rate": 4.90875e-06, + "loss": 0.2347, + "step": 73 + }, + { + "epoch": 0.0875480627033422, + "grad_norm": 2.3564510345458984, + "learning_rate": 4.907500000000001e-06, + "loss": 0.2409, + "step": 74 + }, + { + "epoch": 0.08873114463176575, + "grad_norm": 2.026122808456421, + "learning_rate": 4.90625e-06, + "loss": 0.2462, + "step": 75 + }, + { + "epoch": 0.0899142265601893, + "grad_norm": 1.986960768699646, + "learning_rate": 4.9050000000000005e-06, + "loss": 0.2314, + "step": 76 + }, + { + "epoch": 0.09109730848861283, + "grad_norm": 2.0246565341949463, + "learning_rate": 4.90375e-06, + "loss": 0.2313, + "step": 77 + }, + { + "epoch": 0.09228039041703638, + "grad_norm": 1.7285372018814087, + "learning_rate": 4.902500000000001e-06, + "loss": 0.2273, + "step": 78 + }, + { + "epoch": 0.09346347234545992, + "grad_norm": 2.1053431034088135, + "learning_rate": 4.90125e-06, + "loss": 0.2547, + "step": 79 + }, + { + "epoch": 0.09464655427388347, + "grad_norm": 1.9087393283843994, + "learning_rate": 4.9000000000000005e-06, + "loss": 0.2126, + "step": 80 + }, + { + "epoch": 0.09582963620230701, + "grad_norm": 2.052885055541992, + "learning_rate": 4.89875e-06, + "loss": 0.2496, + "step": 81 + }, + { + "epoch": 0.09701271813073055, + "grad_norm": 2.595057964324951, + "learning_rate": 4.897500000000001e-06, + "loss": 0.2497, + "step": 82 + }, + { + "epoch": 0.0981958000591541, + "grad_norm": 1.749897837638855, + "learning_rate": 4.89625e-06, + "loss": 0.2349, + "step": 83 + }, + { + "epoch": 0.09937888198757763, + "grad_norm": 2.030165195465088, + "learning_rate": 4.8950000000000006e-06, + "loss": 0.254, + "step": 84 + }, + { + "epoch": 0.10056196391600118, + "grad_norm": 2.150524616241455, + "learning_rate": 4.89375e-06, + "loss": 0.2224, + "step": 85 + }, + { + "epoch": 0.10174504584442473, + "grad_norm": 1.9854685068130493, + "learning_rate": 4.8925e-06, + "loss": 0.2354, + "step": 86 + }, + { + "epoch": 0.10292812777284827, + "grad_norm": 1.9026265144348145, + "learning_rate": 4.89125e-06, + "loss": 0.2391, + "step": 87 + }, + { + "epoch": 0.10411120970127181, + "grad_norm": 1.8395593166351318, + "learning_rate": 4.890000000000001e-06, + "loss": 0.2498, + "step": 88 + }, + { + "epoch": 0.10529429162969535, + "grad_norm": 1.9397979974746704, + "learning_rate": 4.88875e-06, + "loss": 0.2377, + "step": 89 + }, + { + "epoch": 0.1064773735581189, + "grad_norm": 1.8917561769485474, + "learning_rate": 4.8875e-06, + "loss": 0.2443, + "step": 90 + }, + { + "epoch": 0.10766045548654245, + "grad_norm": 2.038583517074585, + "learning_rate": 4.88625e-06, + "loss": 0.236, + "step": 91 + }, + { + "epoch": 0.10884353741496598, + "grad_norm": 1.635682463645935, + "learning_rate": 4.885000000000001e-06, + "loss": 0.2216, + "step": 92 + }, + { + "epoch": 0.11002661934338953, + "grad_norm": 1.834957242012024, + "learning_rate": 4.8837500000000005e-06, + "loss": 0.2188, + "step": 93 + }, + { + "epoch": 0.11120970127181307, + "grad_norm": 1.8312525749206543, + "learning_rate": 4.8825e-06, + "loss": 0.2467, + "step": 94 + }, + { + "epoch": 0.11239278320023662, + "grad_norm": 2.0534043312072754, + "learning_rate": 4.88125e-06, + "loss": 0.2501, + "step": 95 + }, + { + "epoch": 0.11357586512866016, + "grad_norm": 1.9417256116867065, + "learning_rate": 4.880000000000001e-06, + "loss": 0.2206, + "step": 96 + }, + { + "epoch": 0.1147589470570837, + "grad_norm": 1.9795541763305664, + "learning_rate": 4.8787500000000005e-06, + "loss": 0.2323, + "step": 97 + }, + { + "epoch": 0.11594202898550725, + "grad_norm": 2.1221625804901123, + "learning_rate": 4.8775e-06, + "loss": 0.223, + "step": 98 + }, + { + "epoch": 0.11712511091393078, + "grad_norm": 1.8060489892959595, + "learning_rate": 4.87625e-06, + "loss": 0.2293, + "step": 99 + }, + { + "epoch": 0.11830819284235433, + "grad_norm": 2.0370259284973145, + "learning_rate": 4.875e-06, + "loss": 0.2126, + "step": 100 + }, + { + "epoch": 0.11949127477077788, + "grad_norm": 1.8419151306152344, + "learning_rate": 4.8737500000000006e-06, + "loss": 0.2279, + "step": 101 + }, + { + "epoch": 0.12067435669920142, + "grad_norm": 1.9191561937332153, + "learning_rate": 4.8725e-06, + "loss": 0.2598, + "step": 102 + }, + { + "epoch": 0.12185743862762496, + "grad_norm": 1.8152592182159424, + "learning_rate": 4.87125e-06, + "loss": 0.2362, + "step": 103 + }, + { + "epoch": 0.12304052055604851, + "grad_norm": 1.8178733587265015, + "learning_rate": 4.87e-06, + "loss": 0.2403, + "step": 104 + }, + { + "epoch": 0.12422360248447205, + "grad_norm": 1.9474964141845703, + "learning_rate": 4.868750000000001e-06, + "loss": 0.2468, + "step": 105 + }, + { + "epoch": 0.1254066844128956, + "grad_norm": 1.9086790084838867, + "learning_rate": 4.8675e-06, + "loss": 0.2086, + "step": 106 + }, + { + "epoch": 0.12658976634131913, + "grad_norm": 2.346707820892334, + "learning_rate": 4.86625e-06, + "loss": 0.2216, + "step": 107 + }, + { + "epoch": 0.12777284826974267, + "grad_norm": 1.738911747932434, + "learning_rate": 4.865e-06, + "loss": 0.2183, + "step": 108 + }, + { + "epoch": 0.12895593019816623, + "grad_norm": 2.1267385482788086, + "learning_rate": 4.863750000000001e-06, + "loss": 0.2455, + "step": 109 + }, + { + "epoch": 0.13013901212658976, + "grad_norm": 1.868361473083496, + "learning_rate": 4.8625000000000005e-06, + "loss": 0.2364, + "step": 110 + }, + { + "epoch": 0.1313220940550133, + "grad_norm": 1.7360605001449585, + "learning_rate": 4.86125e-06, + "loss": 0.2425, + "step": 111 + }, + { + "epoch": 0.13250517598343686, + "grad_norm": 1.9552383422851562, + "learning_rate": 4.86e-06, + "loss": 0.2269, + "step": 112 + }, + { + "epoch": 0.1336882579118604, + "grad_norm": 2.0971555709838867, + "learning_rate": 4.858750000000001e-06, + "loss": 0.2052, + "step": 113 + }, + { + "epoch": 0.13487133984028393, + "grad_norm": 2.4274823665618896, + "learning_rate": 4.8575000000000005e-06, + "loss": 0.2476, + "step": 114 + }, + { + "epoch": 0.1360544217687075, + "grad_norm": 1.9225788116455078, + "learning_rate": 4.85625e-06, + "loss": 0.2427, + "step": 115 + }, + { + "epoch": 0.13723750369713103, + "grad_norm": 2.1598422527313232, + "learning_rate": 4.855e-06, + "loss": 0.2342, + "step": 116 + }, + { + "epoch": 0.13842058562555457, + "grad_norm": 1.8408483266830444, + "learning_rate": 4.853750000000001e-06, + "loss": 0.2399, + "step": 117 + }, + { + "epoch": 0.1396036675539781, + "grad_norm": 2.0003838539123535, + "learning_rate": 4.8525000000000006e-06, + "loss": 0.2397, + "step": 118 + }, + { + "epoch": 0.14078674948240166, + "grad_norm": 1.8507474660873413, + "learning_rate": 4.85125e-06, + "loss": 0.246, + "step": 119 + }, + { + "epoch": 0.1419698314108252, + "grad_norm": 2.0442306995391846, + "learning_rate": 4.85e-06, + "loss": 0.2298, + "step": 120 + }, + { + "epoch": 0.14315291333924873, + "grad_norm": 2.0822994709014893, + "learning_rate": 4.848750000000001e-06, + "loss": 0.2393, + "step": 121 + }, + { + "epoch": 0.1443359952676723, + "grad_norm": 1.8815940618515015, + "learning_rate": 4.847500000000001e-06, + "loss": 0.2374, + "step": 122 + }, + { + "epoch": 0.14551907719609583, + "grad_norm": 1.8023940324783325, + "learning_rate": 4.84625e-06, + "loss": 0.2303, + "step": 123 + }, + { + "epoch": 0.14670215912451937, + "grad_norm": 1.7209413051605225, + "learning_rate": 4.845e-06, + "loss": 0.2349, + "step": 124 + }, + { + "epoch": 0.14788524105294293, + "grad_norm": 1.731852412223816, + "learning_rate": 4.84375e-06, + "loss": 0.2317, + "step": 125 + }, + { + "epoch": 0.14906832298136646, + "grad_norm": 1.9795680046081543, + "learning_rate": 4.842500000000001e-06, + "loss": 0.2353, + "step": 126 + }, + { + "epoch": 0.15025140490979, + "grad_norm": 2.0678229331970215, + "learning_rate": 4.8412500000000004e-06, + "loss": 0.2423, + "step": 127 + }, + { + "epoch": 0.15143448683821353, + "grad_norm": 1.6294100284576416, + "learning_rate": 4.84e-06, + "loss": 0.2347, + "step": 128 + }, + { + "epoch": 0.1526175687666371, + "grad_norm": 1.7438840866088867, + "learning_rate": 4.83875e-06, + "loss": 0.2419, + "step": 129 + }, + { + "epoch": 0.15380065069506063, + "grad_norm": 2.2364532947540283, + "learning_rate": 4.837500000000001e-06, + "loss": 0.2566, + "step": 130 + }, + { + "epoch": 0.15498373262348417, + "grad_norm": 1.8731287717819214, + "learning_rate": 4.8362500000000005e-06, + "loss": 0.2225, + "step": 131 + }, + { + "epoch": 0.15616681455190773, + "grad_norm": 1.8111813068389893, + "learning_rate": 4.835e-06, + "loss": 0.2465, + "step": 132 + }, + { + "epoch": 0.15734989648033126, + "grad_norm": 1.958540916442871, + "learning_rate": 4.83375e-06, + "loss": 0.2203, + "step": 133 + }, + { + "epoch": 0.1585329784087548, + "grad_norm": 1.8410875797271729, + "learning_rate": 4.832500000000001e-06, + "loss": 0.2161, + "step": 134 + }, + { + "epoch": 0.15971606033717836, + "grad_norm": 2.0129263401031494, + "learning_rate": 4.8312500000000005e-06, + "loss": 0.2252, + "step": 135 + }, + { + "epoch": 0.1608991422656019, + "grad_norm": 1.6709232330322266, + "learning_rate": 4.83e-06, + "loss": 0.2361, + "step": 136 + }, + { + "epoch": 0.16208222419402543, + "grad_norm": 1.8482400178909302, + "learning_rate": 4.82875e-06, + "loss": 0.2112, + "step": 137 + }, + { + "epoch": 0.16326530612244897, + "grad_norm": 1.8703190088272095, + "learning_rate": 4.827500000000001e-06, + "loss": 0.2242, + "step": 138 + }, + { + "epoch": 0.16444838805087253, + "grad_norm": 1.6118755340576172, + "learning_rate": 4.826250000000001e-06, + "loss": 0.2262, + "step": 139 + }, + { + "epoch": 0.16563146997929606, + "grad_norm": 1.9231178760528564, + "learning_rate": 4.825e-06, + "loss": 0.2417, + "step": 140 + }, + { + "epoch": 0.1668145519077196, + "grad_norm": 1.9067695140838623, + "learning_rate": 4.82375e-06, + "loss": 0.2498, + "step": 141 + }, + { + "epoch": 0.16799763383614316, + "grad_norm": 2.2337772846221924, + "learning_rate": 4.822500000000001e-06, + "loss": 0.2215, + "step": 142 + }, + { + "epoch": 0.1691807157645667, + "grad_norm": 1.7563021183013916, + "learning_rate": 4.821250000000001e-06, + "loss": 0.2152, + "step": 143 + }, + { + "epoch": 0.17036379769299023, + "grad_norm": 1.7275441884994507, + "learning_rate": 4.8200000000000004e-06, + "loss": 0.2395, + "step": 144 + }, + { + "epoch": 0.1715468796214138, + "grad_norm": 1.9618782997131348, + "learning_rate": 4.81875e-06, + "loss": 0.2569, + "step": 145 + }, + { + "epoch": 0.17272996154983733, + "grad_norm": 1.9048913717269897, + "learning_rate": 4.817500000000001e-06, + "loss": 0.2375, + "step": 146 + }, + { + "epoch": 0.17391304347826086, + "grad_norm": 1.8320634365081787, + "learning_rate": 4.816250000000001e-06, + "loss": 0.2235, + "step": 147 + }, + { + "epoch": 0.1750961254066844, + "grad_norm": 1.8724092245101929, + "learning_rate": 4.8150000000000005e-06, + "loss": 0.2227, + "step": 148 + }, + { + "epoch": 0.17627920733510796, + "grad_norm": 1.9613903760910034, + "learning_rate": 4.81375e-06, + "loss": 0.2334, + "step": 149 + }, + { + "epoch": 0.1774622892635315, + "grad_norm": 1.6121172904968262, + "learning_rate": 4.8125e-06, + "loss": 0.2051, + "step": 150 + }, + { + "epoch": 0.17864537119195503, + "grad_norm": 2.206853151321411, + "learning_rate": 4.811250000000001e-06, + "loss": 0.253, + "step": 151 + }, + { + "epoch": 0.1798284531203786, + "grad_norm": 1.854071855545044, + "learning_rate": 4.8100000000000005e-06, + "loss": 0.236, + "step": 152 + }, + { + "epoch": 0.18101153504880213, + "grad_norm": 2.0291330814361572, + "learning_rate": 4.80875e-06, + "loss": 0.2325, + "step": 153 + }, + { + "epoch": 0.18219461697722567, + "grad_norm": 1.771928071975708, + "learning_rate": 4.8075e-06, + "loss": 0.2192, + "step": 154 + }, + { + "epoch": 0.18337769890564923, + "grad_norm": 1.9444481134414673, + "learning_rate": 4.806250000000001e-06, + "loss": 0.2298, + "step": 155 + }, + { + "epoch": 0.18456078083407276, + "grad_norm": 2.5084357261657715, + "learning_rate": 4.805000000000001e-06, + "loss": 0.2413, + "step": 156 + }, + { + "epoch": 0.1857438627624963, + "grad_norm": 1.6882388591766357, + "learning_rate": 4.80375e-06, + "loss": 0.2167, + "step": 157 + }, + { + "epoch": 0.18692694469091983, + "grad_norm": 2.0198111534118652, + "learning_rate": 4.8025e-06, + "loss": 0.2326, + "step": 158 + }, + { + "epoch": 0.1881100266193434, + "grad_norm": 2.1372225284576416, + "learning_rate": 4.801250000000001e-06, + "loss": 0.2083, + "step": 159 + }, + { + "epoch": 0.18929310854776693, + "grad_norm": 2.214465618133545, + "learning_rate": 4.800000000000001e-06, + "loss": 0.2131, + "step": 160 + }, + { + "epoch": 0.19047619047619047, + "grad_norm": 2.080329179763794, + "learning_rate": 4.7987500000000004e-06, + "loss": 0.2022, + "step": 161 + }, + { + "epoch": 0.19165927240461403, + "grad_norm": 2.0989134311676025, + "learning_rate": 4.7975e-06, + "loss": 0.2383, + "step": 162 + }, + { + "epoch": 0.19284235433303756, + "grad_norm": 1.682469367980957, + "learning_rate": 4.796250000000001e-06, + "loss": 0.215, + "step": 163 + }, + { + "epoch": 0.1940254362614611, + "grad_norm": 1.5963906049728394, + "learning_rate": 4.795e-06, + "loss": 0.2279, + "step": 164 + }, + { + "epoch": 0.19520851818988466, + "grad_norm": 2.0112242698669434, + "learning_rate": 4.7937500000000005e-06, + "loss": 0.2345, + "step": 165 + }, + { + "epoch": 0.1963916001183082, + "grad_norm": 1.5308153629302979, + "learning_rate": 4.7925e-06, + "loss": 0.2224, + "step": 166 + }, + { + "epoch": 0.19757468204673173, + "grad_norm": 1.4766302108764648, + "learning_rate": 4.791250000000001e-06, + "loss": 0.2129, + "step": 167 + }, + { + "epoch": 0.19875776397515527, + "grad_norm": 2.0693554878234863, + "learning_rate": 4.79e-06, + "loss": 0.2344, + "step": 168 + }, + { + "epoch": 0.19994084590357883, + "grad_norm": 1.9057021141052246, + "learning_rate": 4.7887500000000005e-06, + "loss": 0.2523, + "step": 169 + }, + { + "epoch": 0.20112392783200236, + "grad_norm": 1.812928557395935, + "learning_rate": 4.7875e-06, + "loss": 0.2318, + "step": 170 + }, + { + "epoch": 0.2023070097604259, + "grad_norm": 1.6785112619400024, + "learning_rate": 4.786250000000001e-06, + "loss": 0.2078, + "step": 171 + }, + { + "epoch": 0.20349009168884946, + "grad_norm": 1.9895825386047363, + "learning_rate": 4.785e-06, + "loss": 0.2113, + "step": 172 + }, + { + "epoch": 0.204673173617273, + "grad_norm": 2.292640447616577, + "learning_rate": 4.7837500000000006e-06, + "loss": 0.2383, + "step": 173 + }, + { + "epoch": 0.20585625554569653, + "grad_norm": 1.9101717472076416, + "learning_rate": 4.7825e-06, + "loss": 0.2241, + "step": 174 + }, + { + "epoch": 0.2070393374741201, + "grad_norm": 2.0935518741607666, + "learning_rate": 4.781250000000001e-06, + "loss": 0.223, + "step": 175 + }, + { + "epoch": 0.20822241940254363, + "grad_norm": 1.849385380744934, + "learning_rate": 4.78e-06, + "loss": 0.2543, + "step": 176 + }, + { + "epoch": 0.20940550133096716, + "grad_norm": 1.771072506904602, + "learning_rate": 4.778750000000001e-06, + "loss": 0.2047, + "step": 177 + }, + { + "epoch": 0.2105885832593907, + "grad_norm": 1.7452104091644287, + "learning_rate": 4.7775e-06, + "loss": 0.255, + "step": 178 + }, + { + "epoch": 0.21177166518781426, + "grad_norm": 1.7960786819458008, + "learning_rate": 4.77625e-06, + "loss": 0.2139, + "step": 179 + }, + { + "epoch": 0.2129547471162378, + "grad_norm": 1.7545033693313599, + "learning_rate": 4.775e-06, + "loss": 0.2136, + "step": 180 + }, + { + "epoch": 0.21413782904466133, + "grad_norm": 1.925955891609192, + "learning_rate": 4.773750000000001e-06, + "loss": 0.2648, + "step": 181 + }, + { + "epoch": 0.2153209109730849, + "grad_norm": 2.2802045345306396, + "learning_rate": 4.7725000000000005e-06, + "loss": 0.2406, + "step": 182 + }, + { + "epoch": 0.21650399290150843, + "grad_norm": 1.839751124382019, + "learning_rate": 4.77125e-06, + "loss": 0.2354, + "step": 183 + }, + { + "epoch": 0.21768707482993196, + "grad_norm": 1.8213366270065308, + "learning_rate": 4.77e-06, + "loss": 0.2174, + "step": 184 + }, + { + "epoch": 0.21887015675835553, + "grad_norm": 2.1424074172973633, + "learning_rate": 4.768750000000001e-06, + "loss": 0.2299, + "step": 185 + }, + { + "epoch": 0.22005323868677906, + "grad_norm": 2.2080776691436768, + "learning_rate": 4.7675000000000005e-06, + "loss": 0.2582, + "step": 186 + }, + { + "epoch": 0.2212363206152026, + "grad_norm": 1.9374247789382935, + "learning_rate": 4.76625e-06, + "loss": 0.2308, + "step": 187 + }, + { + "epoch": 0.22241940254362613, + "grad_norm": 1.8572752475738525, + "learning_rate": 4.765e-06, + "loss": 0.2224, + "step": 188 + }, + { + "epoch": 0.2236024844720497, + "grad_norm": 1.833842158317566, + "learning_rate": 4.76375e-06, + "loss": 0.209, + "step": 189 + }, + { + "epoch": 0.22478556640047323, + "grad_norm": 1.8651922941207886, + "learning_rate": 4.7625000000000006e-06, + "loss": 0.2416, + "step": 190 + }, + { + "epoch": 0.22596864832889677, + "grad_norm": 1.9629894495010376, + "learning_rate": 4.76125e-06, + "loss": 0.2351, + "step": 191 + }, + { + "epoch": 0.22715173025732033, + "grad_norm": 2.3880834579467773, + "learning_rate": 4.76e-06, + "loss": 0.2277, + "step": 192 + }, + { + "epoch": 0.22833481218574386, + "grad_norm": 1.876425862312317, + "learning_rate": 4.75875e-06, + "loss": 0.2222, + "step": 193 + }, + { + "epoch": 0.2295178941141674, + "grad_norm": 1.9424009323120117, + "learning_rate": 4.757500000000001e-06, + "loss": 0.2084, + "step": 194 + }, + { + "epoch": 0.23070097604259096, + "grad_norm": 1.9733895063400269, + "learning_rate": 4.75625e-06, + "loss": 0.224, + "step": 195 + }, + { + "epoch": 0.2318840579710145, + "grad_norm": 1.9220353364944458, + "learning_rate": 4.755e-06, + "loss": 0.2058, + "step": 196 + }, + { + "epoch": 0.23306713989943803, + "grad_norm": 1.8912702798843384, + "learning_rate": 4.75375e-06, + "loss": 0.2274, + "step": 197 + }, + { + "epoch": 0.23425022182786157, + "grad_norm": 1.6422877311706543, + "learning_rate": 4.752500000000001e-06, + "loss": 0.2098, + "step": 198 + }, + { + "epoch": 0.23543330375628513, + "grad_norm": 1.8763656616210938, + "learning_rate": 4.7512500000000005e-06, + "loss": 0.2422, + "step": 199 + }, + { + "epoch": 0.23661638568470866, + "grad_norm": 1.6136271953582764, + "learning_rate": 4.75e-06, + "loss": 0.2112, + "step": 200 + }, + { + "epoch": 0.2377994676131322, + "grad_norm": 1.6210423707962036, + "learning_rate": 4.74875e-06, + "loss": 0.2135, + "step": 201 + }, + { + "epoch": 0.23898254954155576, + "grad_norm": 1.6347814798355103, + "learning_rate": 4.747500000000001e-06, + "loss": 0.2166, + "step": 202 + }, + { + "epoch": 0.2401656314699793, + "grad_norm": 1.7272768020629883, + "learning_rate": 4.7462500000000005e-06, + "loss": 0.2181, + "step": 203 + }, + { + "epoch": 0.24134871339840283, + "grad_norm": 2.3884904384613037, + "learning_rate": 4.745e-06, + "loss": 0.2093, + "step": 204 + }, + { + "epoch": 0.2425317953268264, + "grad_norm": 1.7122042179107666, + "learning_rate": 4.74375e-06, + "loss": 0.1969, + "step": 205 + }, + { + "epoch": 0.24371487725524993, + "grad_norm": 1.8603320121765137, + "learning_rate": 4.742500000000001e-06, + "loss": 0.2361, + "step": 206 + }, + { + "epoch": 0.24489795918367346, + "grad_norm": 1.915271282196045, + "learning_rate": 4.7412500000000006e-06, + "loss": 0.2111, + "step": 207 + }, + { + "epoch": 0.24608104111209703, + "grad_norm": 2.093996047973633, + "learning_rate": 4.74e-06, + "loss": 0.2246, + "step": 208 + }, + { + "epoch": 0.24726412304052056, + "grad_norm": 1.6151585578918457, + "learning_rate": 4.73875e-06, + "loss": 0.2193, + "step": 209 + }, + { + "epoch": 0.2484472049689441, + "grad_norm": 1.8422338962554932, + "learning_rate": 4.737500000000001e-06, + "loss": 0.2105, + "step": 210 + }, + { + "epoch": 0.24963028689736763, + "grad_norm": 2.2223892211914062, + "learning_rate": 4.736250000000001e-06, + "loss": 0.2263, + "step": 211 + }, + { + "epoch": 0.2508133688257912, + "grad_norm": 2.5265085697174072, + "learning_rate": 4.735e-06, + "loss": 0.2186, + "step": 212 + }, + { + "epoch": 0.25199645075421473, + "grad_norm": 2.063999652862549, + "learning_rate": 4.73375e-06, + "loss": 0.2455, + "step": 213 + }, + { + "epoch": 0.25317953268263826, + "grad_norm": 1.6079577207565308, + "learning_rate": 4.7325e-06, + "loss": 0.2201, + "step": 214 + }, + { + "epoch": 0.2543626146110618, + "grad_norm": 1.671782374382019, + "learning_rate": 4.731250000000001e-06, + "loss": 0.2125, + "step": 215 + }, + { + "epoch": 0.25554569653948533, + "grad_norm": 1.914597749710083, + "learning_rate": 4.7300000000000005e-06, + "loss": 0.221, + "step": 216 + }, + { + "epoch": 0.2567287784679089, + "grad_norm": 1.7564365863800049, + "learning_rate": 4.72875e-06, + "loss": 0.23, + "step": 217 + }, + { + "epoch": 0.25791186039633246, + "grad_norm": 2.4676191806793213, + "learning_rate": 4.7275e-06, + "loss": 0.2021, + "step": 218 + }, + { + "epoch": 0.259094942324756, + "grad_norm": 1.5812536478042603, + "learning_rate": 4.726250000000001e-06, + "loss": 0.1916, + "step": 219 + }, + { + "epoch": 0.26027802425317953, + "grad_norm": 1.5892866849899292, + "learning_rate": 4.7250000000000005e-06, + "loss": 0.2174, + "step": 220 + }, + { + "epoch": 0.26146110618160306, + "grad_norm": 1.6493667364120483, + "learning_rate": 4.72375e-06, + "loss": 0.2036, + "step": 221 + }, + { + "epoch": 0.2626441881100266, + "grad_norm": 1.8415342569351196, + "learning_rate": 4.7225e-06, + "loss": 0.2433, + "step": 222 + }, + { + "epoch": 0.2638272700384502, + "grad_norm": 2.2788166999816895, + "learning_rate": 4.721250000000001e-06, + "loss": 0.2069, + "step": 223 + }, + { + "epoch": 0.2650103519668737, + "grad_norm": 1.7495577335357666, + "learning_rate": 4.7200000000000005e-06, + "loss": 0.2267, + "step": 224 + }, + { + "epoch": 0.26619343389529726, + "grad_norm": 2.159227132797241, + "learning_rate": 4.71875e-06, + "loss": 0.2456, + "step": 225 + }, + { + "epoch": 0.2673765158237208, + "grad_norm": 1.6947096586227417, + "learning_rate": 4.7175e-06, + "loss": 0.2184, + "step": 226 + }, + { + "epoch": 0.26855959775214433, + "grad_norm": 2.0596272945404053, + "learning_rate": 4.716250000000001e-06, + "loss": 0.249, + "step": 227 + }, + { + "epoch": 0.26974267968056787, + "grad_norm": 1.8883509635925293, + "learning_rate": 4.715e-06, + "loss": 0.2157, + "step": 228 + }, + { + "epoch": 0.2709257616089914, + "grad_norm": 1.81239914894104, + "learning_rate": 4.71375e-06, + "loss": 0.2185, + "step": 229 + }, + { + "epoch": 0.272108843537415, + "grad_norm": 1.716597080230713, + "learning_rate": 4.7125e-06, + "loss": 0.2179, + "step": 230 + }, + { + "epoch": 0.2732919254658385, + "grad_norm": 1.643808126449585, + "learning_rate": 4.711250000000001e-06, + "loss": 0.2042, + "step": 231 + }, + { + "epoch": 0.27447500739426206, + "grad_norm": 1.8664461374282837, + "learning_rate": 4.71e-06, + "loss": 0.2237, + "step": 232 + }, + { + "epoch": 0.2756580893226856, + "grad_norm": 1.9388753175735474, + "learning_rate": 4.7087500000000004e-06, + "loss": 0.2152, + "step": 233 + }, + { + "epoch": 0.27684117125110913, + "grad_norm": 1.9597885608673096, + "learning_rate": 4.7075e-06, + "loss": 0.2128, + "step": 234 + }, + { + "epoch": 0.27802425317953267, + "grad_norm": 1.7251286506652832, + "learning_rate": 4.706250000000001e-06, + "loss": 0.2081, + "step": 235 + }, + { + "epoch": 0.2792073351079562, + "grad_norm": 1.7104822397232056, + "learning_rate": 4.705e-06, + "loss": 0.2193, + "step": 236 + }, + { + "epoch": 0.2803904170363798, + "grad_norm": 1.8473135232925415, + "learning_rate": 4.7037500000000005e-06, + "loss": 0.2172, + "step": 237 + }, + { + "epoch": 0.2815734989648033, + "grad_norm": 1.7523866891860962, + "learning_rate": 4.7025e-06, + "loss": 0.2273, + "step": 238 + }, + { + "epoch": 0.28275658089322686, + "grad_norm": 1.7974885702133179, + "learning_rate": 4.701250000000001e-06, + "loss": 0.2087, + "step": 239 + }, + { + "epoch": 0.2839396628216504, + "grad_norm": 1.8637504577636719, + "learning_rate": 4.7e-06, + "loss": 0.203, + "step": 240 + }, + { + "epoch": 0.28512274475007393, + "grad_norm": 1.6129474639892578, + "learning_rate": 4.6987500000000005e-06, + "loss": 0.2272, + "step": 241 + }, + { + "epoch": 0.28630582667849747, + "grad_norm": 1.7749543190002441, + "learning_rate": 4.6975e-06, + "loss": 0.2202, + "step": 242 + }, + { + "epoch": 0.28748890860692106, + "grad_norm": 1.5690652132034302, + "learning_rate": 4.69625e-06, + "loss": 0.203, + "step": 243 + }, + { + "epoch": 0.2886719905353446, + "grad_norm": 1.6161121129989624, + "learning_rate": 4.695e-06, + "loss": 0.2066, + "step": 244 + }, + { + "epoch": 0.2898550724637681, + "grad_norm": 2.0193095207214355, + "learning_rate": 4.693750000000001e-06, + "loss": 0.2311, + "step": 245 + }, + { + "epoch": 0.29103815439219166, + "grad_norm": 1.7465118169784546, + "learning_rate": 4.6925e-06, + "loss": 0.2368, + "step": 246 + }, + { + "epoch": 0.2922212363206152, + "grad_norm": 1.782921552658081, + "learning_rate": 4.69125e-06, + "loss": 0.226, + "step": 247 + }, + { + "epoch": 0.29340431824903873, + "grad_norm": 2.059583902359009, + "learning_rate": 4.69e-06, + "loss": 0.2284, + "step": 248 + }, + { + "epoch": 0.29458740017746227, + "grad_norm": 1.7890956401824951, + "learning_rate": 4.688750000000001e-06, + "loss": 0.2081, + "step": 249 + }, + { + "epoch": 0.29577048210588586, + "grad_norm": 1.9331074953079224, + "learning_rate": 4.6875000000000004e-06, + "loss": 0.2252, + "step": 250 + }, + { + "epoch": 0.2969535640343094, + "grad_norm": 1.836992621421814, + "learning_rate": 4.68625e-06, + "loss": 0.1971, + "step": 251 + }, + { + "epoch": 0.2981366459627329, + "grad_norm": 1.8596571683883667, + "learning_rate": 4.685000000000001e-06, + "loss": 0.2321, + "step": 252 + }, + { + "epoch": 0.29931972789115646, + "grad_norm": 2.118957281112671, + "learning_rate": 4.68375e-06, + "loss": 0.2086, + "step": 253 + }, + { + "epoch": 0.30050280981958, + "grad_norm": 1.8454577922821045, + "learning_rate": 4.6825000000000005e-06, + "loss": 0.2292, + "step": 254 + }, + { + "epoch": 0.30168589174800353, + "grad_norm": 1.4728995561599731, + "learning_rate": 4.68125e-06, + "loss": 0.1977, + "step": 255 + }, + { + "epoch": 0.30286897367642707, + "grad_norm": 1.6834067106246948, + "learning_rate": 4.680000000000001e-06, + "loss": 0.2323, + "step": 256 + }, + { + "epoch": 0.30405205560485066, + "grad_norm": 1.6254894733428955, + "learning_rate": 4.67875e-06, + "loss": 0.2355, + "step": 257 + }, + { + "epoch": 0.3052351375332742, + "grad_norm": 1.7059401273727417, + "learning_rate": 4.6775000000000005e-06, + "loss": 0.2042, + "step": 258 + }, + { + "epoch": 0.3064182194616977, + "grad_norm": 1.8304659128189087, + "learning_rate": 4.67625e-06, + "loss": 0.2368, + "step": 259 + }, + { + "epoch": 0.30760130139012126, + "grad_norm": 1.5870059728622437, + "learning_rate": 4.675000000000001e-06, + "loss": 0.2367, + "step": 260 + }, + { + "epoch": 0.3087843833185448, + "grad_norm": 1.6900827884674072, + "learning_rate": 4.67375e-06, + "loss": 0.2345, + "step": 261 + }, + { + "epoch": 0.30996746524696833, + "grad_norm": 1.6485545635223389, + "learning_rate": 4.672500000000001e-06, + "loss": 0.2046, + "step": 262 + }, + { + "epoch": 0.3111505471753919, + "grad_norm": 1.7679157257080078, + "learning_rate": 4.67125e-06, + "loss": 0.2332, + "step": 263 + }, + { + "epoch": 0.31233362910381546, + "grad_norm": 1.676711082458496, + "learning_rate": 4.670000000000001e-06, + "loss": 0.2149, + "step": 264 + }, + { + "epoch": 0.313516711032239, + "grad_norm": 1.7871328592300415, + "learning_rate": 4.66875e-06, + "loss": 0.2234, + "step": 265 + }, + { + "epoch": 0.3146997929606625, + "grad_norm": 1.9631577730178833, + "learning_rate": 4.667500000000001e-06, + "loss": 0.2087, + "step": 266 + }, + { + "epoch": 0.31588287488908606, + "grad_norm": 1.6207168102264404, + "learning_rate": 4.6662500000000004e-06, + "loss": 0.2284, + "step": 267 + }, + { + "epoch": 0.3170659568175096, + "grad_norm": 1.7508774995803833, + "learning_rate": 4.665e-06, + "loss": 0.1945, + "step": 268 + }, + { + "epoch": 0.31824903874593313, + "grad_norm": 1.8421528339385986, + "learning_rate": 4.66375e-06, + "loss": 0.2178, + "step": 269 + }, + { + "epoch": 0.3194321206743567, + "grad_norm": 2.0203309059143066, + "learning_rate": 4.662500000000001e-06, + "loss": 0.2204, + "step": 270 + }, + { + "epoch": 0.32061520260278026, + "grad_norm": 2.0495989322662354, + "learning_rate": 4.6612500000000005e-06, + "loss": 0.2349, + "step": 271 + }, + { + "epoch": 0.3217982845312038, + "grad_norm": 1.5367748737335205, + "learning_rate": 4.66e-06, + "loss": 0.218, + "step": 272 + }, + { + "epoch": 0.32298136645962733, + "grad_norm": 1.9693152904510498, + "learning_rate": 4.65875e-06, + "loss": 0.225, + "step": 273 + }, + { + "epoch": 0.32416444838805086, + "grad_norm": 1.571284532546997, + "learning_rate": 4.657500000000001e-06, + "loss": 0.1901, + "step": 274 + }, + { + "epoch": 0.3253475303164744, + "grad_norm": 1.801742434501648, + "learning_rate": 4.6562500000000005e-06, + "loss": 0.215, + "step": 275 + }, + { + "epoch": 0.32653061224489793, + "grad_norm": 1.6383665800094604, + "learning_rate": 4.655e-06, + "loss": 0.2043, + "step": 276 + }, + { + "epoch": 0.3277136941733215, + "grad_norm": 1.9774034023284912, + "learning_rate": 4.65375e-06, + "loss": 0.2118, + "step": 277 + }, + { + "epoch": 0.32889677610174506, + "grad_norm": 2.110351324081421, + "learning_rate": 4.652500000000001e-06, + "loss": 0.2127, + "step": 278 + }, + { + "epoch": 0.3300798580301686, + "grad_norm": 1.775883436203003, + "learning_rate": 4.6512500000000006e-06, + "loss": 0.2312, + "step": 279 + }, + { + "epoch": 0.33126293995859213, + "grad_norm": 1.783628225326538, + "learning_rate": 4.65e-06, + "loss": 0.2134, + "step": 280 + }, + { + "epoch": 0.33244602188701566, + "grad_norm": 1.6660100221633911, + "learning_rate": 4.64875e-06, + "loss": 0.2025, + "step": 281 + }, + { + "epoch": 0.3336291038154392, + "grad_norm": 1.7838362455368042, + "learning_rate": 4.6475e-06, + "loss": 0.1952, + "step": 282 + }, + { + "epoch": 0.3348121857438628, + "grad_norm": 2.0005300045013428, + "learning_rate": 4.646250000000001e-06, + "loss": 0.2235, + "step": 283 + }, + { + "epoch": 0.3359952676722863, + "grad_norm": 1.7513573169708252, + "learning_rate": 4.645e-06, + "loss": 0.2216, + "step": 284 + }, + { + "epoch": 0.33717834960070986, + "grad_norm": 1.5646618604660034, + "learning_rate": 4.64375e-06, + "loss": 0.2438, + "step": 285 + }, + { + "epoch": 0.3383614315291334, + "grad_norm": 2.242877244949341, + "learning_rate": 4.6425e-06, + "loss": 0.2357, + "step": 286 + }, + { + "epoch": 0.33954451345755693, + "grad_norm": 1.6213918924331665, + "learning_rate": 4.641250000000001e-06, + "loss": 0.2018, + "step": 287 + }, + { + "epoch": 0.34072759538598046, + "grad_norm": 1.6741118431091309, + "learning_rate": 4.6400000000000005e-06, + "loss": 0.2045, + "step": 288 + }, + { + "epoch": 0.341910677314404, + "grad_norm": 1.8860020637512207, + "learning_rate": 4.63875e-06, + "loss": 0.2311, + "step": 289 + }, + { + "epoch": 0.3430937592428276, + "grad_norm": 1.8140908479690552, + "learning_rate": 4.6375e-06, + "loss": 0.2082, + "step": 290 + }, + { + "epoch": 0.3442768411712511, + "grad_norm": 2.069559335708618, + "learning_rate": 4.636250000000001e-06, + "loss": 0.217, + "step": 291 + }, + { + "epoch": 0.34545992309967466, + "grad_norm": 1.786759614944458, + "learning_rate": 4.6350000000000005e-06, + "loss": 0.2459, + "step": 292 + }, + { + "epoch": 0.3466430050280982, + "grad_norm": 2.1142795085906982, + "learning_rate": 4.63375e-06, + "loss": 0.2182, + "step": 293 + }, + { + "epoch": 0.34782608695652173, + "grad_norm": 1.65966796875, + "learning_rate": 4.6325e-06, + "loss": 0.226, + "step": 294 + }, + { + "epoch": 0.34900916888494526, + "grad_norm": 1.7002227306365967, + "learning_rate": 4.631250000000001e-06, + "loss": 0.2051, + "step": 295 + }, + { + "epoch": 0.3501922508133688, + "grad_norm": 1.5609714984893799, + "learning_rate": 4.6300000000000006e-06, + "loss": 0.2292, + "step": 296 + }, + { + "epoch": 0.3513753327417924, + "grad_norm": 1.6661500930786133, + "learning_rate": 4.62875e-06, + "loss": 0.2145, + "step": 297 + }, + { + "epoch": 0.3525584146702159, + "grad_norm": 1.8161113262176514, + "learning_rate": 4.6275e-06, + "loss": 0.202, + "step": 298 + }, + { + "epoch": 0.35374149659863946, + "grad_norm": 1.5403625965118408, + "learning_rate": 4.626250000000001e-06, + "loss": 0.2127, + "step": 299 + }, + { + "epoch": 0.354924578527063, + "grad_norm": 2.010000467300415, + "learning_rate": 4.625000000000001e-06, + "loss": 0.2203, + "step": 300 + }, + { + "epoch": 0.35610766045548653, + "grad_norm": 1.7578142881393433, + "learning_rate": 4.62375e-06, + "loss": 0.2154, + "step": 301 + }, + { + "epoch": 0.35729074238391006, + "grad_norm": 1.9667818546295166, + "learning_rate": 4.6225e-06, + "loss": 0.2444, + "step": 302 + }, + { + "epoch": 0.35847382431233366, + "grad_norm": 1.7061426639556885, + "learning_rate": 4.621250000000001e-06, + "loss": 0.209, + "step": 303 + }, + { + "epoch": 0.3596569062407572, + "grad_norm": 1.7255232334136963, + "learning_rate": 4.620000000000001e-06, + "loss": 0.1999, + "step": 304 + }, + { + "epoch": 0.3608399881691807, + "grad_norm": 2.0622966289520264, + "learning_rate": 4.6187500000000005e-06, + "loss": 0.2303, + "step": 305 + }, + { + "epoch": 0.36202307009760426, + "grad_norm": 1.749866008758545, + "learning_rate": 4.6175e-06, + "loss": 0.1898, + "step": 306 + }, + { + "epoch": 0.3632061520260278, + "grad_norm": 1.9503586292266846, + "learning_rate": 4.61625e-06, + "loss": 0.2177, + "step": 307 + }, + { + "epoch": 0.36438923395445133, + "grad_norm": 1.7881333827972412, + "learning_rate": 4.615000000000001e-06, + "loss": 0.2067, + "step": 308 + }, + { + "epoch": 0.36557231588287487, + "grad_norm": 1.5984553098678589, + "learning_rate": 4.6137500000000005e-06, + "loss": 0.2059, + "step": 309 + }, + { + "epoch": 0.36675539781129846, + "grad_norm": 1.6921051740646362, + "learning_rate": 4.6125e-06, + "loss": 0.1969, + "step": 310 + }, + { + "epoch": 0.367938479739722, + "grad_norm": 1.6874794960021973, + "learning_rate": 4.61125e-06, + "loss": 0.2099, + "step": 311 + }, + { + "epoch": 0.3691215616681455, + "grad_norm": 1.8225075006484985, + "learning_rate": 4.610000000000001e-06, + "loss": 0.2056, + "step": 312 + }, + { + "epoch": 0.37030464359656906, + "grad_norm": 1.9512572288513184, + "learning_rate": 4.6087500000000006e-06, + "loss": 0.2055, + "step": 313 + }, + { + "epoch": 0.3714877255249926, + "grad_norm": 1.6682723760604858, + "learning_rate": 4.6075e-06, + "loss": 0.2427, + "step": 314 + }, + { + "epoch": 0.37267080745341613, + "grad_norm": 1.918190836906433, + "learning_rate": 4.60625e-06, + "loss": 0.2056, + "step": 315 + }, + { + "epoch": 0.37385388938183967, + "grad_norm": 1.7301883697509766, + "learning_rate": 4.605000000000001e-06, + "loss": 0.2281, + "step": 316 + }, + { + "epoch": 0.37503697131026326, + "grad_norm": 1.6691696643829346, + "learning_rate": 4.60375e-06, + "loss": 0.2156, + "step": 317 + }, + { + "epoch": 0.3762200532386868, + "grad_norm": 1.6302396059036255, + "learning_rate": 4.6025e-06, + "loss": 0.2065, + "step": 318 + }, + { + "epoch": 0.3774031351671103, + "grad_norm": 1.833674669265747, + "learning_rate": 4.60125e-06, + "loss": 0.229, + "step": 319 + }, + { + "epoch": 0.37858621709553386, + "grad_norm": 1.5501904487609863, + "learning_rate": 4.600000000000001e-06, + "loss": 0.2171, + "step": 320 + }, + { + "epoch": 0.3797692990239574, + "grad_norm": 1.7691535949707031, + "learning_rate": 4.59875e-06, + "loss": 0.2003, + "step": 321 + }, + { + "epoch": 0.38095238095238093, + "grad_norm": 1.8329675197601318, + "learning_rate": 4.5975000000000005e-06, + "loss": 0.2309, + "step": 322 + }, + { + "epoch": 0.3821354628808045, + "grad_norm": 1.7573652267456055, + "learning_rate": 4.59625e-06, + "loss": 0.2148, + "step": 323 + }, + { + "epoch": 0.38331854480922806, + "grad_norm": 1.7804630994796753, + "learning_rate": 4.595000000000001e-06, + "loss": 0.2054, + "step": 324 + }, + { + "epoch": 0.3845016267376516, + "grad_norm": 1.6134300231933594, + "learning_rate": 4.59375e-06, + "loss": 0.1997, + "step": 325 + }, + { + "epoch": 0.3856847086660751, + "grad_norm": 1.727796196937561, + "learning_rate": 4.5925000000000005e-06, + "loss": 0.2216, + "step": 326 + }, + { + "epoch": 0.38686779059449866, + "grad_norm": 1.699853539466858, + "learning_rate": 4.59125e-06, + "loss": 0.2169, + "step": 327 + }, + { + "epoch": 0.3880508725229222, + "grad_norm": 1.673011064529419, + "learning_rate": 4.590000000000001e-06, + "loss": 0.2073, + "step": 328 + }, + { + "epoch": 0.38923395445134573, + "grad_norm": 1.7647662162780762, + "learning_rate": 4.58875e-06, + "loss": 0.2317, + "step": 329 + }, + { + "epoch": 0.3904170363797693, + "grad_norm": 2.058245897293091, + "learning_rate": 4.5875000000000005e-06, + "loss": 0.2107, + "step": 330 + }, + { + "epoch": 0.39160011830819286, + "grad_norm": 2.0578060150146484, + "learning_rate": 4.58625e-06, + "loss": 0.2032, + "step": 331 + }, + { + "epoch": 0.3927832002366164, + "grad_norm": 1.7520545721054077, + "learning_rate": 4.585e-06, + "loss": 0.2133, + "step": 332 + }, + { + "epoch": 0.3939662821650399, + "grad_norm": 1.805719256401062, + "learning_rate": 4.58375e-06, + "loss": 0.2061, + "step": 333 + }, + { + "epoch": 0.39514936409346346, + "grad_norm": 1.4834145307540894, + "learning_rate": 4.582500000000001e-06, + "loss": 0.2097, + "step": 334 + }, + { + "epoch": 0.396332446021887, + "grad_norm": 1.9067150354385376, + "learning_rate": 4.58125e-06, + "loss": 0.2074, + "step": 335 + }, + { + "epoch": 0.39751552795031053, + "grad_norm": 1.6015737056732178, + "learning_rate": 4.58e-06, + "loss": 0.2213, + "step": 336 + }, + { + "epoch": 0.3986986098787341, + "grad_norm": 1.7946031093597412, + "learning_rate": 4.57875e-06, + "loss": 0.2143, + "step": 337 + }, + { + "epoch": 0.39988169180715766, + "grad_norm": 1.5006310939788818, + "learning_rate": 4.577500000000001e-06, + "loss": 0.2021, + "step": 338 + }, + { + "epoch": 0.4010647737355812, + "grad_norm": 1.6249268054962158, + "learning_rate": 4.5762500000000004e-06, + "loss": 0.2061, + "step": 339 + }, + { + "epoch": 0.4022478556640047, + "grad_norm": 1.7842872142791748, + "learning_rate": 4.575e-06, + "loss": 0.2205, + "step": 340 + }, + { + "epoch": 0.40343093759242826, + "grad_norm": 1.849157691001892, + "learning_rate": 4.57375e-06, + "loss": 0.2026, + "step": 341 + }, + { + "epoch": 0.4046140195208518, + "grad_norm": 1.9391974210739136, + "learning_rate": 4.572500000000001e-06, + "loss": 0.2184, + "step": 342 + }, + { + "epoch": 0.4057971014492754, + "grad_norm": 1.933650016784668, + "learning_rate": 4.5712500000000005e-06, + "loss": 0.2281, + "step": 343 + }, + { + "epoch": 0.4069801833776989, + "grad_norm": 1.7541342973709106, + "learning_rate": 4.57e-06, + "loss": 0.1919, + "step": 344 + }, + { + "epoch": 0.40816326530612246, + "grad_norm": 1.7624635696411133, + "learning_rate": 4.56875e-06, + "loss": 0.2279, + "step": 345 + }, + { + "epoch": 0.409346347234546, + "grad_norm": 1.3936728239059448, + "learning_rate": 4.5675e-06, + "loss": 0.206, + "step": 346 + }, + { + "epoch": 0.41052942916296953, + "grad_norm": 1.860927700996399, + "learning_rate": 4.5662500000000005e-06, + "loss": 0.2371, + "step": 347 + }, + { + "epoch": 0.41171251109139306, + "grad_norm": 1.6294583082199097, + "learning_rate": 4.565e-06, + "loss": 0.2058, + "step": 348 + }, + { + "epoch": 0.4128955930198166, + "grad_norm": 1.4565120935440063, + "learning_rate": 4.56375e-06, + "loss": 0.2003, + "step": 349 + }, + { + "epoch": 0.4140786749482402, + "grad_norm": 1.7280256748199463, + "learning_rate": 4.5625e-06, + "loss": 0.2129, + "step": 350 + }, + { + "epoch": 0.4152617568766637, + "grad_norm": 1.8258018493652344, + "learning_rate": 4.561250000000001e-06, + "loss": 0.2303, + "step": 351 + }, + { + "epoch": 0.41644483880508726, + "grad_norm": 1.644449234008789, + "learning_rate": 4.56e-06, + "loss": 0.23, + "step": 352 + }, + { + "epoch": 0.4176279207335108, + "grad_norm": 1.6997029781341553, + "learning_rate": 4.55875e-06, + "loss": 0.2356, + "step": 353 + }, + { + "epoch": 0.41881100266193433, + "grad_norm": 2.184889793395996, + "learning_rate": 4.5575e-06, + "loss": 0.2078, + "step": 354 + }, + { + "epoch": 0.41999408459035786, + "grad_norm": 1.5398186445236206, + "learning_rate": 4.556250000000001e-06, + "loss": 0.2095, + "step": 355 + }, + { + "epoch": 0.4211771665187814, + "grad_norm": 1.7635611295700073, + "learning_rate": 4.5550000000000004e-06, + "loss": 0.2174, + "step": 356 + }, + { + "epoch": 0.422360248447205, + "grad_norm": 1.5516245365142822, + "learning_rate": 4.55375e-06, + "loss": 0.1984, + "step": 357 + }, + { + "epoch": 0.4235433303756285, + "grad_norm": 1.7472184896469116, + "learning_rate": 4.5525e-06, + "loss": 0.2112, + "step": 358 + }, + { + "epoch": 0.42472641230405206, + "grad_norm": 1.7334010601043701, + "learning_rate": 4.551250000000001e-06, + "loss": 0.2143, + "step": 359 + }, + { + "epoch": 0.4259094942324756, + "grad_norm": 1.5825953483581543, + "learning_rate": 4.5500000000000005e-06, + "loss": 0.2106, + "step": 360 + }, + { + "epoch": 0.42709257616089913, + "grad_norm": 1.4335472583770752, + "learning_rate": 4.54875e-06, + "loss": 0.2065, + "step": 361 + }, + { + "epoch": 0.42827565808932266, + "grad_norm": 1.914963722229004, + "learning_rate": 4.5475e-06, + "loss": 0.2398, + "step": 362 + }, + { + "epoch": 0.42945874001774625, + "grad_norm": 1.6975761651992798, + "learning_rate": 4.546250000000001e-06, + "loss": 0.21, + "step": 363 + }, + { + "epoch": 0.4306418219461698, + "grad_norm": 1.5699235200881958, + "learning_rate": 4.5450000000000005e-06, + "loss": 0.204, + "step": 364 + }, + { + "epoch": 0.4318249038745933, + "grad_norm": 1.6318728923797607, + "learning_rate": 4.54375e-06, + "loss": 0.2034, + "step": 365 + }, + { + "epoch": 0.43300798580301686, + "grad_norm": 1.8014485836029053, + "learning_rate": 4.5425e-06, + "loss": 0.2169, + "step": 366 + }, + { + "epoch": 0.4341910677314404, + "grad_norm": 1.6654741764068604, + "learning_rate": 4.541250000000001e-06, + "loss": 0.2242, + "step": 367 + }, + { + "epoch": 0.43537414965986393, + "grad_norm": 1.9408992528915405, + "learning_rate": 4.540000000000001e-06, + "loss": 0.2189, + "step": 368 + }, + { + "epoch": 0.43655723158828746, + "grad_norm": 1.5695315599441528, + "learning_rate": 4.53875e-06, + "loss": 0.1971, + "step": 369 + }, + { + "epoch": 0.43774031351671105, + "grad_norm": 1.936196208000183, + "learning_rate": 4.5375e-06, + "loss": 0.215, + "step": 370 + }, + { + "epoch": 0.4389233954451346, + "grad_norm": 1.5989937782287598, + "learning_rate": 4.53625e-06, + "loss": 0.212, + "step": 371 + }, + { + "epoch": 0.4401064773735581, + "grad_norm": 1.7612370252609253, + "learning_rate": 4.535000000000001e-06, + "loss": 0.1892, + "step": 372 + }, + { + "epoch": 0.44128955930198166, + "grad_norm": 2.814040422439575, + "learning_rate": 4.5337500000000004e-06, + "loss": 0.1988, + "step": 373 + }, + { + "epoch": 0.4424726412304052, + "grad_norm": 1.6128261089324951, + "learning_rate": 4.5325e-06, + "loss": 0.1968, + "step": 374 + }, + { + "epoch": 0.44365572315882873, + "grad_norm": 1.6829733848571777, + "learning_rate": 4.53125e-06, + "loss": 0.1983, + "step": 375 + }, + { + "epoch": 0.44483880508725226, + "grad_norm": 1.7525848150253296, + "learning_rate": 4.530000000000001e-06, + "loss": 0.2236, + "step": 376 + }, + { + "epoch": 0.44602188701567586, + "grad_norm": 1.5992765426635742, + "learning_rate": 4.5287500000000005e-06, + "loss": 0.1862, + "step": 377 + }, + { + "epoch": 0.4472049689440994, + "grad_norm": 1.7822481393814087, + "learning_rate": 4.5275e-06, + "loss": 0.2106, + "step": 378 + }, + { + "epoch": 0.4483880508725229, + "grad_norm": 1.827082633972168, + "learning_rate": 4.52625e-06, + "loss": 0.2171, + "step": 379 + }, + { + "epoch": 0.44957113280094646, + "grad_norm": 1.8720605373382568, + "learning_rate": 4.525000000000001e-06, + "loss": 0.1804, + "step": 380 + }, + { + "epoch": 0.45075421472937, + "grad_norm": 1.6681894063949585, + "learning_rate": 4.5237500000000005e-06, + "loss": 0.1964, + "step": 381 + }, + { + "epoch": 0.45193729665779353, + "grad_norm": 1.5934771299362183, + "learning_rate": 4.5225e-06, + "loss": 0.1933, + "step": 382 + }, + { + "epoch": 0.4531203785862171, + "grad_norm": 1.6890853643417358, + "learning_rate": 4.52125e-06, + "loss": 0.1983, + "step": 383 + }, + { + "epoch": 0.45430346051464066, + "grad_norm": 1.5356639623641968, + "learning_rate": 4.520000000000001e-06, + "loss": 0.2037, + "step": 384 + }, + { + "epoch": 0.4554865424430642, + "grad_norm": 1.8971006870269775, + "learning_rate": 4.5187500000000006e-06, + "loss": 0.2208, + "step": 385 + }, + { + "epoch": 0.4566696243714877, + "grad_norm": 1.688204288482666, + "learning_rate": 4.5175e-06, + "loss": 0.2211, + "step": 386 + }, + { + "epoch": 0.45785270629991126, + "grad_norm": 1.6857317686080933, + "learning_rate": 4.51625e-06, + "loss": 0.2125, + "step": 387 + }, + { + "epoch": 0.4590357882283348, + "grad_norm": 1.7091715335845947, + "learning_rate": 4.515000000000001e-06, + "loss": 0.2286, + "step": 388 + }, + { + "epoch": 0.46021887015675833, + "grad_norm": 1.5651905536651611, + "learning_rate": 4.513750000000001e-06, + "loss": 0.2095, + "step": 389 + }, + { + "epoch": 0.4614019520851819, + "grad_norm": 1.79058039188385, + "learning_rate": 4.5125e-06, + "loss": 0.2333, + "step": 390 + }, + { + "epoch": 0.46258503401360546, + "grad_norm": 1.6704179048538208, + "learning_rate": 4.51125e-06, + "loss": 0.2, + "step": 391 + }, + { + "epoch": 0.463768115942029, + "grad_norm": 1.5676157474517822, + "learning_rate": 4.510000000000001e-06, + "loss": 0.1921, + "step": 392 + }, + { + "epoch": 0.4649511978704525, + "grad_norm": 1.719386339187622, + "learning_rate": 4.508750000000001e-06, + "loss": 0.224, + "step": 393 + }, + { + "epoch": 0.46613427979887606, + "grad_norm": 1.467307209968567, + "learning_rate": 4.5075000000000005e-06, + "loss": 0.2147, + "step": 394 + }, + { + "epoch": 0.4673173617272996, + "grad_norm": 1.7842071056365967, + "learning_rate": 4.50625e-06, + "loss": 0.226, + "step": 395 + }, + { + "epoch": 0.46850044365572313, + "grad_norm": 1.8227238655090332, + "learning_rate": 4.505e-06, + "loss": 0.2027, + "step": 396 + }, + { + "epoch": 0.4696835255841467, + "grad_norm": 1.6178356409072876, + "learning_rate": 4.503750000000001e-06, + "loss": 0.2026, + "step": 397 + }, + { + "epoch": 0.47086660751257026, + "grad_norm": 1.7770763635635376, + "learning_rate": 4.5025000000000005e-06, + "loss": 0.2495, + "step": 398 + }, + { + "epoch": 0.4720496894409938, + "grad_norm": 1.7434635162353516, + "learning_rate": 4.50125e-06, + "loss": 0.2118, + "step": 399 + }, + { + "epoch": 0.4732327713694173, + "grad_norm": 1.7318893671035767, + "learning_rate": 4.5e-06, + "loss": 0.2226, + "step": 400 + }, + { + "epoch": 0.47441585329784086, + "grad_norm": 1.8993306159973145, + "learning_rate": 4.498750000000001e-06, + "loss": 0.2056, + "step": 401 + }, + { + "epoch": 0.4755989352262644, + "grad_norm": 1.3445546627044678, + "learning_rate": 4.4975000000000006e-06, + "loss": 0.2175, + "step": 402 + }, + { + "epoch": 0.476782017154688, + "grad_norm": 1.6262346506118774, + "learning_rate": 4.49625e-06, + "loss": 0.2209, + "step": 403 + }, + { + "epoch": 0.4779650990831115, + "grad_norm": 1.5847078561782837, + "learning_rate": 4.495e-06, + "loss": 0.2155, + "step": 404 + }, + { + "epoch": 0.47914818101153506, + "grad_norm": 1.6557165384292603, + "learning_rate": 4.493750000000001e-06, + "loss": 0.2039, + "step": 405 + }, + { + "epoch": 0.4803312629399586, + "grad_norm": 1.7115685939788818, + "learning_rate": 4.492500000000001e-06, + "loss": 0.2332, + "step": 406 + }, + { + "epoch": 0.4815143448683821, + "grad_norm": 2.348205089569092, + "learning_rate": 4.49125e-06, + "loss": 0.2476, + "step": 407 + }, + { + "epoch": 0.48269742679680566, + "grad_norm": 1.6784801483154297, + "learning_rate": 4.49e-06, + "loss": 0.2295, + "step": 408 + }, + { + "epoch": 0.4838805087252292, + "grad_norm": 1.6796497106552124, + "learning_rate": 4.488750000000001e-06, + "loss": 0.2179, + "step": 409 + }, + { + "epoch": 0.4850635906536528, + "grad_norm": 1.6482985019683838, + "learning_rate": 4.4875e-06, + "loss": 0.2136, + "step": 410 + }, + { + "epoch": 0.4862466725820763, + "grad_norm": 1.8032575845718384, + "learning_rate": 4.4862500000000005e-06, + "loss": 0.2189, + "step": 411 + }, + { + "epoch": 0.48742975451049986, + "grad_norm": 1.6482560634613037, + "learning_rate": 4.485e-06, + "loss": 0.2093, + "step": 412 + }, + { + "epoch": 0.4886128364389234, + "grad_norm": 1.7061456441879272, + "learning_rate": 4.483750000000001e-06, + "loss": 0.1986, + "step": 413 + }, + { + "epoch": 0.4897959183673469, + "grad_norm": 1.7199705839157104, + "learning_rate": 4.4825e-06, + "loss": 0.239, + "step": 414 + }, + { + "epoch": 0.49097900029577046, + "grad_norm": 1.5953779220581055, + "learning_rate": 4.4812500000000005e-06, + "loss": 0.2163, + "step": 415 + }, + { + "epoch": 0.49216208222419405, + "grad_norm": 1.6749153137207031, + "learning_rate": 4.48e-06, + "loss": 0.2283, + "step": 416 + }, + { + "epoch": 0.4933451641526176, + "grad_norm": 1.7338147163391113, + "learning_rate": 4.478750000000001e-06, + "loss": 0.2166, + "step": 417 + }, + { + "epoch": 0.4945282460810411, + "grad_norm": 1.5903236865997314, + "learning_rate": 4.4775e-06, + "loss": 0.2361, + "step": 418 + }, + { + "epoch": 0.49571132800946466, + "grad_norm": 1.6387348175048828, + "learning_rate": 4.4762500000000006e-06, + "loss": 0.201, + "step": 419 + }, + { + "epoch": 0.4968944099378882, + "grad_norm": 1.6347439289093018, + "learning_rate": 4.475e-06, + "loss": 0.2105, + "step": 420 + }, + { + "epoch": 0.49807749186631173, + "grad_norm": 1.7331531047821045, + "learning_rate": 4.473750000000001e-06, + "loss": 0.2108, + "step": 421 + }, + { + "epoch": 0.49926057379473526, + "grad_norm": 2.0043935775756836, + "learning_rate": 4.4725e-06, + "loss": 0.2244, + "step": 422 + }, + { + "epoch": 0.5004436557231589, + "grad_norm": 1.4389022588729858, + "learning_rate": 4.471250000000001e-06, + "loss": 0.2123, + "step": 423 + }, + { + "epoch": 0.5016267376515824, + "grad_norm": 1.7766342163085938, + "learning_rate": 4.47e-06, + "loss": 0.1937, + "step": 424 + }, + { + "epoch": 0.5028098195800059, + "grad_norm": 1.5891697406768799, + "learning_rate": 4.46875e-06, + "loss": 0.2246, + "step": 425 + }, + { + "epoch": 0.5039929015084295, + "grad_norm": 1.7131683826446533, + "learning_rate": 4.4675e-06, + "loss": 0.2044, + "step": 426 + }, + { + "epoch": 0.505175983436853, + "grad_norm": 1.6454190015792847, + "learning_rate": 4.466250000000001e-06, + "loss": 0.2094, + "step": 427 + }, + { + "epoch": 0.5063590653652765, + "grad_norm": 1.5807616710662842, + "learning_rate": 4.4650000000000004e-06, + "loss": 0.2073, + "step": 428 + }, + { + "epoch": 0.5075421472937001, + "grad_norm": 1.5518710613250732, + "learning_rate": 4.46375e-06, + "loss": 0.2126, + "step": 429 + }, + { + "epoch": 0.5087252292221236, + "grad_norm": 1.5535545349121094, + "learning_rate": 4.4625e-06, + "loss": 0.2162, + "step": 430 + }, + { + "epoch": 0.5099083111505471, + "grad_norm": 1.9821901321411133, + "learning_rate": 4.461250000000001e-06, + "loss": 0.2261, + "step": 431 + }, + { + "epoch": 0.5110913930789707, + "grad_norm": 1.7361701726913452, + "learning_rate": 4.4600000000000005e-06, + "loss": 0.2086, + "step": 432 + }, + { + "epoch": 0.5122744750073943, + "grad_norm": 1.683637261390686, + "learning_rate": 4.45875e-06, + "loss": 0.2175, + "step": 433 + }, + { + "epoch": 0.5134575569358178, + "grad_norm": 1.6871002912521362, + "learning_rate": 4.4575e-06, + "loss": 0.2069, + "step": 434 + }, + { + "epoch": 0.5146406388642414, + "grad_norm": 1.5806211233139038, + "learning_rate": 4.45625e-06, + "loss": 0.2121, + "step": 435 + }, + { + "epoch": 0.5158237207926649, + "grad_norm": 1.4649739265441895, + "learning_rate": 4.4550000000000005e-06, + "loss": 0.2072, + "step": 436 + }, + { + "epoch": 0.5170068027210885, + "grad_norm": 1.8463490009307861, + "learning_rate": 4.45375e-06, + "loss": 0.2238, + "step": 437 + }, + { + "epoch": 0.518189884649512, + "grad_norm": 1.4759910106658936, + "learning_rate": 4.4525e-06, + "loss": 0.2017, + "step": 438 + }, + { + "epoch": 0.5193729665779355, + "grad_norm": 1.6824644804000854, + "learning_rate": 4.45125e-06, + "loss": 0.2149, + "step": 439 + }, + { + "epoch": 0.5205560485063591, + "grad_norm": 2.995065689086914, + "learning_rate": 4.450000000000001e-06, + "loss": 0.2135, + "step": 440 + }, + { + "epoch": 0.5217391304347826, + "grad_norm": 1.481467366218567, + "learning_rate": 4.44875e-06, + "loss": 0.1848, + "step": 441 + }, + { + "epoch": 0.5229222123632061, + "grad_norm": 1.5626577138900757, + "learning_rate": 4.4475e-06, + "loss": 0.2122, + "step": 442 + }, + { + "epoch": 0.5241052942916297, + "grad_norm": 1.439121961593628, + "learning_rate": 4.44625e-06, + "loss": 0.2029, + "step": 443 + }, + { + "epoch": 0.5252883762200532, + "grad_norm": 1.6135458946228027, + "learning_rate": 4.445000000000001e-06, + "loss": 0.2266, + "step": 444 + }, + { + "epoch": 0.5264714581484767, + "grad_norm": 1.9590203762054443, + "learning_rate": 4.4437500000000004e-06, + "loss": 0.2209, + "step": 445 + }, + { + "epoch": 0.5276545400769004, + "grad_norm": 1.9306215047836304, + "learning_rate": 4.4425e-06, + "loss": 0.217, + "step": 446 + }, + { + "epoch": 0.5288376220053239, + "grad_norm": 2.0270955562591553, + "learning_rate": 4.44125e-06, + "loss": 0.2039, + "step": 447 + }, + { + "epoch": 0.5300207039337475, + "grad_norm": 1.7091397047042847, + "learning_rate": 4.440000000000001e-06, + "loss": 0.2055, + "step": 448 + }, + { + "epoch": 0.531203785862171, + "grad_norm": 1.8021570444107056, + "learning_rate": 4.4387500000000005e-06, + "loss": 0.2309, + "step": 449 + }, + { + "epoch": 0.5323868677905945, + "grad_norm": 1.5575273036956787, + "learning_rate": 4.4375e-06, + "loss": 0.1852, + "step": 450 + }, + { + "epoch": 0.533569949719018, + "grad_norm": 1.7770799398422241, + "learning_rate": 4.43625e-06, + "loss": 0.1941, + "step": 451 + }, + { + "epoch": 0.5347530316474416, + "grad_norm": 1.7122217416763306, + "learning_rate": 4.435000000000001e-06, + "loss": 0.1998, + "step": 452 + }, + { + "epoch": 0.5359361135758651, + "grad_norm": 1.657188892364502, + "learning_rate": 4.4337500000000005e-06, + "loss": 0.2027, + "step": 453 + }, + { + "epoch": 0.5371191955042887, + "grad_norm": 1.7258437871932983, + "learning_rate": 4.4325e-06, + "loss": 0.2428, + "step": 454 + }, + { + "epoch": 0.5383022774327122, + "grad_norm": 1.6258562803268433, + "learning_rate": 4.43125e-06, + "loss": 0.2036, + "step": 455 + }, + { + "epoch": 0.5394853593611357, + "grad_norm": 1.6597267389297485, + "learning_rate": 4.430000000000001e-06, + "loss": 0.2011, + "step": 456 + }, + { + "epoch": 0.5406684412895593, + "grad_norm": 1.5381706953048706, + "learning_rate": 4.428750000000001e-06, + "loss": 0.2266, + "step": 457 + }, + { + "epoch": 0.5418515232179828, + "grad_norm": 1.6866801977157593, + "learning_rate": 4.4275e-06, + "loss": 0.2063, + "step": 458 + }, + { + "epoch": 0.5430346051464063, + "grad_norm": 1.5725903511047363, + "learning_rate": 4.42625e-06, + "loss": 0.214, + "step": 459 + }, + { + "epoch": 0.54421768707483, + "grad_norm": 2.0236740112304688, + "learning_rate": 4.425e-06, + "loss": 0.2222, + "step": 460 + }, + { + "epoch": 0.5454007690032535, + "grad_norm": 1.7172229290008545, + "learning_rate": 4.423750000000001e-06, + "loss": 0.2058, + "step": 461 + }, + { + "epoch": 0.546583850931677, + "grad_norm": 1.7374191284179688, + "learning_rate": 4.4225000000000004e-06, + "loss": 0.2147, + "step": 462 + }, + { + "epoch": 0.5477669328601006, + "grad_norm": 1.895710825920105, + "learning_rate": 4.42125e-06, + "loss": 0.2238, + "step": 463 + }, + { + "epoch": 0.5489500147885241, + "grad_norm": 1.4556119441986084, + "learning_rate": 4.42e-06, + "loss": 0.2117, + "step": 464 + }, + { + "epoch": 0.5501330967169477, + "grad_norm": 1.7223838567733765, + "learning_rate": 4.418750000000001e-06, + "loss": 0.2495, + "step": 465 + }, + { + "epoch": 0.5513161786453712, + "grad_norm": 1.61219322681427, + "learning_rate": 4.4175000000000005e-06, + "loss": 0.2043, + "step": 466 + }, + { + "epoch": 0.5524992605737947, + "grad_norm": 1.5616453886032104, + "learning_rate": 4.41625e-06, + "loss": 0.2135, + "step": 467 + }, + { + "epoch": 0.5536823425022183, + "grad_norm": 1.783302903175354, + "learning_rate": 4.415e-06, + "loss": 0.2216, + "step": 468 + }, + { + "epoch": 0.5548654244306418, + "grad_norm": 1.5496917963027954, + "learning_rate": 4.413750000000001e-06, + "loss": 0.1874, + "step": 469 + }, + { + "epoch": 0.5560485063590653, + "grad_norm": 1.615372657775879, + "learning_rate": 4.4125000000000005e-06, + "loss": 0.2112, + "step": 470 + }, + { + "epoch": 0.5572315882874889, + "grad_norm": 1.6950315237045288, + "learning_rate": 4.41125e-06, + "loss": 0.1994, + "step": 471 + }, + { + "epoch": 0.5584146702159124, + "grad_norm": 1.518828272819519, + "learning_rate": 4.41e-06, + "loss": 0.1993, + "step": 472 + }, + { + "epoch": 0.559597752144336, + "grad_norm": 2.3577628135681152, + "learning_rate": 4.408750000000001e-06, + "loss": 0.2435, + "step": 473 + }, + { + "epoch": 0.5607808340727596, + "grad_norm": 1.615455150604248, + "learning_rate": 4.4075e-06, + "loss": 0.2104, + "step": 474 + }, + { + "epoch": 0.5619639160011831, + "grad_norm": 1.6070345640182495, + "learning_rate": 4.40625e-06, + "loss": 0.2006, + "step": 475 + }, + { + "epoch": 0.5631469979296067, + "grad_norm": 1.4901756048202515, + "learning_rate": 4.405e-06, + "loss": 0.1902, + "step": 476 + }, + { + "epoch": 0.5643300798580302, + "grad_norm": 1.4536210298538208, + "learning_rate": 4.403750000000001e-06, + "loss": 0.1827, + "step": 477 + }, + { + "epoch": 0.5655131617864537, + "grad_norm": 1.6499384641647339, + "learning_rate": 4.4025e-06, + "loss": 0.2103, + "step": 478 + }, + { + "epoch": 0.5666962437148773, + "grad_norm": 1.509630799293518, + "learning_rate": 4.4012500000000004e-06, + "loss": 0.1984, + "step": 479 + }, + { + "epoch": 0.5678793256433008, + "grad_norm": 1.7133687734603882, + "learning_rate": 4.4e-06, + "loss": 0.2099, + "step": 480 + }, + { + "epoch": 0.5690624075717243, + "grad_norm": 1.53462815284729, + "learning_rate": 4.398750000000001e-06, + "loss": 0.2207, + "step": 481 + }, + { + "epoch": 0.5702454895001479, + "grad_norm": 1.5538537502288818, + "learning_rate": 4.3975e-06, + "loss": 0.2121, + "step": 482 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 1.5832204818725586, + "learning_rate": 4.3962500000000005e-06, + "loss": 0.2249, + "step": 483 + }, + { + "epoch": 0.5726116533569949, + "grad_norm": 1.7422429323196411, + "learning_rate": 4.395e-06, + "loss": 0.2189, + "step": 484 + }, + { + "epoch": 0.5737947352854185, + "grad_norm": 1.6034152507781982, + "learning_rate": 4.393750000000001e-06, + "loss": 0.1837, + "step": 485 + }, + { + "epoch": 0.5749778172138421, + "grad_norm": 1.7433987855911255, + "learning_rate": 4.3925e-06, + "loss": 0.2053, + "step": 486 + }, + { + "epoch": 0.5761608991422656, + "grad_norm": 1.4724169969558716, + "learning_rate": 4.3912500000000005e-06, + "loss": 0.2225, + "step": 487 + }, + { + "epoch": 0.5773439810706892, + "grad_norm": 1.6079198122024536, + "learning_rate": 4.39e-06, + "loss": 0.2012, + "step": 488 + }, + { + "epoch": 0.5785270629991127, + "grad_norm": 1.968277931213379, + "learning_rate": 4.38875e-06, + "loss": 0.2142, + "step": 489 + }, + { + "epoch": 0.5797101449275363, + "grad_norm": 1.7785066366195679, + "learning_rate": 4.3875e-06, + "loss": 0.1991, + "step": 490 + }, + { + "epoch": 0.5808932268559598, + "grad_norm": 1.8791481256484985, + "learning_rate": 4.3862500000000006e-06, + "loss": 0.2042, + "step": 491 + }, + { + "epoch": 0.5820763087843833, + "grad_norm": 1.691999912261963, + "learning_rate": 4.385e-06, + "loss": 0.2066, + "step": 492 + }, + { + "epoch": 0.5832593907128069, + "grad_norm": 1.5374629497528076, + "learning_rate": 4.38375e-06, + "loss": 0.1889, + "step": 493 + }, + { + "epoch": 0.5844424726412304, + "grad_norm": 1.5023765563964844, + "learning_rate": 4.3825e-06, + "loss": 0.2163, + "step": 494 + }, + { + "epoch": 0.5856255545696539, + "grad_norm": 1.683051586151123, + "learning_rate": 4.381250000000001e-06, + "loss": 0.2127, + "step": 495 + }, + { + "epoch": 0.5868086364980775, + "grad_norm": 1.5823239088058472, + "learning_rate": 4.38e-06, + "loss": 0.2094, + "step": 496 + }, + { + "epoch": 0.587991718426501, + "grad_norm": 1.4280823469161987, + "learning_rate": 4.37875e-06, + "loss": 0.1903, + "step": 497 + }, + { + "epoch": 0.5891748003549245, + "grad_norm": 1.578560471534729, + "learning_rate": 4.3775e-06, + "loss": 0.226, + "step": 498 + }, + { + "epoch": 0.5903578822833481, + "grad_norm": 2.0102100372314453, + "learning_rate": 4.37625e-06, + "loss": 0.2278, + "step": 499 + }, + { + "epoch": 0.5915409642117717, + "grad_norm": 1.6240367889404297, + "learning_rate": 4.3750000000000005e-06, + "loss": 0.2103, + "step": 500 + }, + { + "epoch": 0.5927240461401952, + "grad_norm": 1.9728202819824219, + "learning_rate": 4.37375e-06, + "loss": 0.1947, + "step": 501 + }, + { + "epoch": 0.5939071280686188, + "grad_norm": 1.6145192384719849, + "learning_rate": 4.372500000000001e-06, + "loss": 0.2297, + "step": 502 + }, + { + "epoch": 0.5950902099970423, + "grad_norm": 1.6598877906799316, + "learning_rate": 4.37125e-06, + "loss": 0.2185, + "step": 503 + }, + { + "epoch": 0.5962732919254659, + "grad_norm": 2.293994665145874, + "learning_rate": 4.3700000000000005e-06, + "loss": 0.1986, + "step": 504 + }, + { + "epoch": 0.5974563738538894, + "grad_norm": 1.6783469915390015, + "learning_rate": 4.36875e-06, + "loss": 0.2264, + "step": 505 + }, + { + "epoch": 0.5986394557823129, + "grad_norm": 1.482409119606018, + "learning_rate": 4.367500000000001e-06, + "loss": 0.2109, + "step": 506 + }, + { + "epoch": 0.5998225377107365, + "grad_norm": 1.8433934450149536, + "learning_rate": 4.36625e-06, + "loss": 0.2363, + "step": 507 + }, + { + "epoch": 0.60100561963916, + "grad_norm": 1.5579134225845337, + "learning_rate": 4.3650000000000006e-06, + "loss": 0.2184, + "step": 508 + }, + { + "epoch": 0.6021887015675835, + "grad_norm": 1.572428822517395, + "learning_rate": 4.36375e-06, + "loss": 0.1958, + "step": 509 + }, + { + "epoch": 0.6033717834960071, + "grad_norm": 1.5066767930984497, + "learning_rate": 4.362500000000001e-06, + "loss": 0.2086, + "step": 510 + }, + { + "epoch": 0.6045548654244306, + "grad_norm": 1.5888288021087646, + "learning_rate": 4.36125e-06, + "loss": 0.2176, + "step": 511 + }, + { + "epoch": 0.6057379473528541, + "grad_norm": 1.5221326351165771, + "learning_rate": 4.360000000000001e-06, + "loss": 0.2475, + "step": 512 + }, + { + "epoch": 0.6069210292812778, + "grad_norm": 1.5014455318450928, + "learning_rate": 4.35875e-06, + "loss": 0.1951, + "step": 513 + }, + { + "epoch": 0.6081041112097013, + "grad_norm": 1.631729245185852, + "learning_rate": 4.3575e-06, + "loss": 0.2081, + "step": 514 + }, + { + "epoch": 0.6092871931381248, + "grad_norm": 1.6843829154968262, + "learning_rate": 4.35625e-06, + "loss": 0.1924, + "step": 515 + }, + { + "epoch": 0.6104702750665484, + "grad_norm": 1.7378168106079102, + "learning_rate": 4.355000000000001e-06, + "loss": 0.2048, + "step": 516 + }, + { + "epoch": 0.6116533569949719, + "grad_norm": 2.033291816711426, + "learning_rate": 4.3537500000000005e-06, + "loss": 0.2255, + "step": 517 + }, + { + "epoch": 0.6128364389233955, + "grad_norm": 1.7761667966842651, + "learning_rate": 4.3525e-06, + "loss": 0.2241, + "step": 518 + }, + { + "epoch": 0.614019520851819, + "grad_norm": 1.4135394096374512, + "learning_rate": 4.35125e-06, + "loss": 0.2043, + "step": 519 + }, + { + "epoch": 0.6152026027802425, + "grad_norm": 1.639792561531067, + "learning_rate": 4.350000000000001e-06, + "loss": 0.1851, + "step": 520 + }, + { + "epoch": 0.6163856847086661, + "grad_norm": 2.0756006240844727, + "learning_rate": 4.3487500000000005e-06, + "loss": 0.2201, + "step": 521 + }, + { + "epoch": 0.6175687666370896, + "grad_norm": 1.8910590410232544, + "learning_rate": 4.3475e-06, + "loss": 0.1961, + "step": 522 + }, + { + "epoch": 0.6187518485655131, + "grad_norm": 1.559146761894226, + "learning_rate": 4.34625e-06, + "loss": 0.1975, + "step": 523 + }, + { + "epoch": 0.6199349304939367, + "grad_norm": 1.6624624729156494, + "learning_rate": 4.345000000000001e-06, + "loss": 0.2188, + "step": 524 + }, + { + "epoch": 0.6211180124223602, + "grad_norm": 1.5368016958236694, + "learning_rate": 4.3437500000000006e-06, + "loss": 0.2337, + "step": 525 + }, + { + "epoch": 0.6223010943507838, + "grad_norm": 1.5993680953979492, + "learning_rate": 4.3425e-06, + "loss": 0.2132, + "step": 526 + }, + { + "epoch": 0.6234841762792074, + "grad_norm": 1.7875285148620605, + "learning_rate": 4.34125e-06, + "loss": 0.2002, + "step": 527 + }, + { + "epoch": 0.6246672582076309, + "grad_norm": 1.596529483795166, + "learning_rate": 4.34e-06, + "loss": 0.2088, + "step": 528 + }, + { + "epoch": 0.6258503401360545, + "grad_norm": 1.6564373970031738, + "learning_rate": 4.338750000000001e-06, + "loss": 0.1986, + "step": 529 + }, + { + "epoch": 0.627033422064478, + "grad_norm": 1.525248646736145, + "learning_rate": 4.3375e-06, + "loss": 0.2048, + "step": 530 + }, + { + "epoch": 0.6282165039929015, + "grad_norm": 1.7580149173736572, + "learning_rate": 4.33625e-06, + "loss": 0.1974, + "step": 531 + }, + { + "epoch": 0.629399585921325, + "grad_norm": 1.8663511276245117, + "learning_rate": 4.335e-06, + "loss": 0.2394, + "step": 532 + }, + { + "epoch": 0.6305826678497486, + "grad_norm": 1.62790846824646, + "learning_rate": 4.333750000000001e-06, + "loss": 0.2331, + "step": 533 + }, + { + "epoch": 0.6317657497781721, + "grad_norm": 1.779070258140564, + "learning_rate": 4.3325000000000004e-06, + "loss": 0.2137, + "step": 534 + }, + { + "epoch": 0.6329488317065957, + "grad_norm": 1.7710734605789185, + "learning_rate": 4.33125e-06, + "loss": 0.2077, + "step": 535 + }, + { + "epoch": 0.6341319136350192, + "grad_norm": 1.9664084911346436, + "learning_rate": 4.33e-06, + "loss": 0.2068, + "step": 536 + }, + { + "epoch": 0.6353149955634427, + "grad_norm": 1.531204342842102, + "learning_rate": 4.328750000000001e-06, + "loss": 0.2063, + "step": 537 + }, + { + "epoch": 0.6364980774918663, + "grad_norm": 1.5048524141311646, + "learning_rate": 4.3275000000000005e-06, + "loss": 0.203, + "step": 538 + }, + { + "epoch": 0.6376811594202898, + "grad_norm": 1.5382282733917236, + "learning_rate": 4.32625e-06, + "loss": 0.2191, + "step": 539 + }, + { + "epoch": 0.6388642413487134, + "grad_norm": 1.8054078817367554, + "learning_rate": 4.325e-06, + "loss": 0.2056, + "step": 540 + }, + { + "epoch": 0.640047323277137, + "grad_norm": 1.687674880027771, + "learning_rate": 4.323750000000001e-06, + "loss": 0.2064, + "step": 541 + }, + { + "epoch": 0.6412304052055605, + "grad_norm": 1.46994149684906, + "learning_rate": 4.3225000000000005e-06, + "loss": 0.1995, + "step": 542 + }, + { + "epoch": 0.642413487133984, + "grad_norm": 1.5751527547836304, + "learning_rate": 4.32125e-06, + "loss": 0.2166, + "step": 543 + }, + { + "epoch": 0.6435965690624076, + "grad_norm": 1.8317450284957886, + "learning_rate": 4.32e-06, + "loss": 0.2466, + "step": 544 + }, + { + "epoch": 0.6447796509908311, + "grad_norm": 1.7101622819900513, + "learning_rate": 4.318750000000001e-06, + "loss": 0.1918, + "step": 545 + }, + { + "epoch": 0.6459627329192547, + "grad_norm": 1.7415885925292969, + "learning_rate": 4.317500000000001e-06, + "loss": 0.2088, + "step": 546 + }, + { + "epoch": 0.6471458148476782, + "grad_norm": 1.7366551160812378, + "learning_rate": 4.31625e-06, + "loss": 0.2116, + "step": 547 + }, + { + "epoch": 0.6483288967761017, + "grad_norm": 1.8025892972946167, + "learning_rate": 4.315e-06, + "loss": 0.1987, + "step": 548 + }, + { + "epoch": 0.6495119787045253, + "grad_norm": 1.4474846124649048, + "learning_rate": 4.313750000000001e-06, + "loss": 0.1918, + "step": 549 + }, + { + "epoch": 0.6506950606329488, + "grad_norm": 1.4484525918960571, + "learning_rate": 4.312500000000001e-06, + "loss": 0.1925, + "step": 550 + }, + { + "epoch": 0.6518781425613723, + "grad_norm": 1.751229166984558, + "learning_rate": 4.3112500000000004e-06, + "loss": 0.2009, + "step": 551 + }, + { + "epoch": 0.6530612244897959, + "grad_norm": 1.5041435956954956, + "learning_rate": 4.31e-06, + "loss": 0.2044, + "step": 552 + }, + { + "epoch": 0.6542443064182195, + "grad_norm": 1.5611358880996704, + "learning_rate": 4.30875e-06, + "loss": 0.1859, + "step": 553 + }, + { + "epoch": 0.655427388346643, + "grad_norm": 2.226494073867798, + "learning_rate": 4.307500000000001e-06, + "loss": 0.2255, + "step": 554 + }, + { + "epoch": 0.6566104702750666, + "grad_norm": 1.728208303451538, + "learning_rate": 4.3062500000000005e-06, + "loss": 0.1863, + "step": 555 + }, + { + "epoch": 0.6577935522034901, + "grad_norm": 1.744835615158081, + "learning_rate": 4.305e-06, + "loss": 0.2182, + "step": 556 + }, + { + "epoch": 0.6589766341319137, + "grad_norm": 1.6456149816513062, + "learning_rate": 4.30375e-06, + "loss": 0.1905, + "step": 557 + }, + { + "epoch": 0.6601597160603372, + "grad_norm": 1.4476972818374634, + "learning_rate": 4.302500000000001e-06, + "loss": 0.2129, + "step": 558 + }, + { + "epoch": 0.6613427979887607, + "grad_norm": 1.708628535270691, + "learning_rate": 4.3012500000000005e-06, + "loss": 0.1886, + "step": 559 + }, + { + "epoch": 0.6625258799171843, + "grad_norm": 1.5854579210281372, + "learning_rate": 4.3e-06, + "loss": 0.1758, + "step": 560 + }, + { + "epoch": 0.6637089618456078, + "grad_norm": 1.320114016532898, + "learning_rate": 4.29875e-06, + "loss": 0.1815, + "step": 561 + }, + { + "epoch": 0.6648920437740313, + "grad_norm": 2.1049113273620605, + "learning_rate": 4.297500000000001e-06, + "loss": 0.1988, + "step": 562 + }, + { + "epoch": 0.6660751257024549, + "grad_norm": 1.402346134185791, + "learning_rate": 4.29625e-06, + "loss": 0.1985, + "step": 563 + }, + { + "epoch": 0.6672582076308784, + "grad_norm": 1.6612980365753174, + "learning_rate": 4.295e-06, + "loss": 0.1982, + "step": 564 + }, + { + "epoch": 0.6684412895593019, + "grad_norm": 1.4486169815063477, + "learning_rate": 4.29375e-06, + "loss": 0.203, + "step": 565 + }, + { + "epoch": 0.6696243714877256, + "grad_norm": 1.7966748476028442, + "learning_rate": 4.292500000000001e-06, + "loss": 0.1975, + "step": 566 + }, + { + "epoch": 0.6708074534161491, + "grad_norm": 1.4396536350250244, + "learning_rate": 4.29125e-06, + "loss": 0.1921, + "step": 567 + }, + { + "epoch": 0.6719905353445726, + "grad_norm": 1.6205607652664185, + "learning_rate": 4.2900000000000004e-06, + "loss": 0.2114, + "step": 568 + }, + { + "epoch": 0.6731736172729962, + "grad_norm": 1.6514461040496826, + "learning_rate": 4.28875e-06, + "loss": 0.2187, + "step": 569 + }, + { + "epoch": 0.6743566992014197, + "grad_norm": 1.6655731201171875, + "learning_rate": 4.287500000000001e-06, + "loss": 0.213, + "step": 570 + }, + { + "epoch": 0.6755397811298433, + "grad_norm": 1.7774158716201782, + "learning_rate": 4.28625e-06, + "loss": 0.1955, + "step": 571 + }, + { + "epoch": 0.6767228630582668, + "grad_norm": 1.7506986856460571, + "learning_rate": 4.2850000000000005e-06, + "loss": 0.2106, + "step": 572 + }, + { + "epoch": 0.6779059449866903, + "grad_norm": 1.5096219778060913, + "learning_rate": 4.28375e-06, + "loss": 0.1921, + "step": 573 + }, + { + "epoch": 0.6790890269151139, + "grad_norm": 1.780593991279602, + "learning_rate": 4.282500000000001e-06, + "loss": 0.2004, + "step": 574 + }, + { + "epoch": 0.6802721088435374, + "grad_norm": 2.072110652923584, + "learning_rate": 4.28125e-06, + "loss": 0.2036, + "step": 575 + }, + { + "epoch": 0.6814551907719609, + "grad_norm": 1.3319354057312012, + "learning_rate": 4.2800000000000005e-06, + "loss": 0.2069, + "step": 576 + }, + { + "epoch": 0.6826382727003845, + "grad_norm": 1.8231171369552612, + "learning_rate": 4.27875e-06, + "loss": 0.2092, + "step": 577 + }, + { + "epoch": 0.683821354628808, + "grad_norm": 1.4618875980377197, + "learning_rate": 4.2775e-06, + "loss": 0.2035, + "step": 578 + }, + { + "epoch": 0.6850044365572315, + "grad_norm": 1.7364439964294434, + "learning_rate": 4.27625e-06, + "loss": 0.2078, + "step": 579 + }, + { + "epoch": 0.6861875184856552, + "grad_norm": 1.625386357307434, + "learning_rate": 4.2750000000000006e-06, + "loss": 0.2143, + "step": 580 + }, + { + "epoch": 0.6873706004140787, + "grad_norm": 1.9209976196289062, + "learning_rate": 4.27375e-06, + "loss": 0.1963, + "step": 581 + }, + { + "epoch": 0.6885536823425022, + "grad_norm": 1.6223750114440918, + "learning_rate": 4.2725e-06, + "loss": 0.2364, + "step": 582 + }, + { + "epoch": 0.6897367642709258, + "grad_norm": 1.997733235359192, + "learning_rate": 4.27125e-06, + "loss": 0.1998, + "step": 583 + }, + { + "epoch": 0.6909198461993493, + "grad_norm": 1.7136197090148926, + "learning_rate": 4.270000000000001e-06, + "loss": 0.1961, + "step": 584 + }, + { + "epoch": 0.6921029281277729, + "grad_norm": 1.4122158288955688, + "learning_rate": 4.26875e-06, + "loss": 0.218, + "step": 585 + }, + { + "epoch": 0.6932860100561964, + "grad_norm": 1.7144594192504883, + "learning_rate": 4.2675e-06, + "loss": 0.2137, + "step": 586 + }, + { + "epoch": 0.6944690919846199, + "grad_norm": 1.5826365947723389, + "learning_rate": 4.26625e-06, + "loss": 0.1911, + "step": 587 + }, + { + "epoch": 0.6956521739130435, + "grad_norm": 1.776910662651062, + "learning_rate": 4.265000000000001e-06, + "loss": 0.2056, + "step": 588 + }, + { + "epoch": 0.696835255841467, + "grad_norm": 1.7742866277694702, + "learning_rate": 4.2637500000000005e-06, + "loss": 0.1879, + "step": 589 + }, + { + "epoch": 0.6980183377698905, + "grad_norm": 1.7628719806671143, + "learning_rate": 4.2625e-06, + "loss": 0.1991, + "step": 590 + }, + { + "epoch": 0.6992014196983141, + "grad_norm": 1.9701482057571411, + "learning_rate": 4.26125e-06, + "loss": 0.2151, + "step": 591 + }, + { + "epoch": 0.7003845016267376, + "grad_norm": 1.6897889375686646, + "learning_rate": 4.26e-06, + "loss": 0.2165, + "step": 592 + }, + { + "epoch": 0.7015675835551612, + "grad_norm": 1.790535807609558, + "learning_rate": 4.2587500000000005e-06, + "loss": 0.206, + "step": 593 + }, + { + "epoch": 0.7027506654835848, + "grad_norm": 1.6010583639144897, + "learning_rate": 4.2575e-06, + "loss": 0.2053, + "step": 594 + }, + { + "epoch": 0.7039337474120083, + "grad_norm": 1.6660845279693604, + "learning_rate": 4.25625e-06, + "loss": 0.2011, + "step": 595 + }, + { + "epoch": 0.7051168293404318, + "grad_norm": 1.6146034002304077, + "learning_rate": 4.255e-06, + "loss": 0.2146, + "step": 596 + }, + { + "epoch": 0.7062999112688554, + "grad_norm": 1.7786117792129517, + "learning_rate": 4.2537500000000006e-06, + "loss": 0.2198, + "step": 597 + }, + { + "epoch": 0.7074829931972789, + "grad_norm": 2.0103049278259277, + "learning_rate": 4.2525e-06, + "loss": 0.2182, + "step": 598 + }, + { + "epoch": 0.7086660751257025, + "grad_norm": 1.7767647504806519, + "learning_rate": 4.25125e-06, + "loss": 0.1991, + "step": 599 + }, + { + "epoch": 0.709849157054126, + "grad_norm": 1.6881451606750488, + "learning_rate": 4.25e-06, + "loss": 0.2077, + "step": 600 + }, + { + "epoch": 0.7110322389825495, + "grad_norm": 1.7919851541519165, + "learning_rate": 4.248750000000001e-06, + "loss": 0.216, + "step": 601 + }, + { + "epoch": 0.7122153209109731, + "grad_norm": 1.5410152673721313, + "learning_rate": 4.2475e-06, + "loss": 0.1943, + "step": 602 + }, + { + "epoch": 0.7133984028393966, + "grad_norm": 1.7030224800109863, + "learning_rate": 4.24625e-06, + "loss": 0.2196, + "step": 603 + }, + { + "epoch": 0.7145814847678201, + "grad_norm": 1.7657229900360107, + "learning_rate": 4.245e-06, + "loss": 0.2064, + "step": 604 + }, + { + "epoch": 0.7157645666962437, + "grad_norm": 1.5063420534133911, + "learning_rate": 4.243750000000001e-06, + "loss": 0.2235, + "step": 605 + }, + { + "epoch": 0.7169476486246673, + "grad_norm": 1.5164815187454224, + "learning_rate": 4.2425000000000005e-06, + "loss": 0.2083, + "step": 606 + }, + { + "epoch": 0.7181307305530908, + "grad_norm": 1.4675016403198242, + "learning_rate": 4.24125e-06, + "loss": 0.201, + "step": 607 + }, + { + "epoch": 0.7193138124815144, + "grad_norm": 1.4602693319320679, + "learning_rate": 4.24e-06, + "loss": 0.2047, + "step": 608 + }, + { + "epoch": 0.7204968944099379, + "grad_norm": 1.4604649543762207, + "learning_rate": 4.238750000000001e-06, + "loss": 0.2114, + "step": 609 + }, + { + "epoch": 0.7216799763383615, + "grad_norm": 1.6961243152618408, + "learning_rate": 4.2375000000000005e-06, + "loss": 0.2031, + "step": 610 + }, + { + "epoch": 0.722863058266785, + "grad_norm": 1.627652883529663, + "learning_rate": 4.23625e-06, + "loss": 0.1946, + "step": 611 + }, + { + "epoch": 0.7240461401952085, + "grad_norm": 1.620342493057251, + "learning_rate": 4.235e-06, + "loss": 0.2002, + "step": 612 + }, + { + "epoch": 0.7252292221236321, + "grad_norm": 1.6774191856384277, + "learning_rate": 4.233750000000001e-06, + "loss": 0.228, + "step": 613 + }, + { + "epoch": 0.7264123040520556, + "grad_norm": 1.731034278869629, + "learning_rate": 4.2325000000000006e-06, + "loss": 0.2193, + "step": 614 + }, + { + "epoch": 0.7275953859804791, + "grad_norm": 1.7032495737075806, + "learning_rate": 4.23125e-06, + "loss": 0.2113, + "step": 615 + }, + { + "epoch": 0.7287784679089027, + "grad_norm": 1.5963670015335083, + "learning_rate": 4.23e-06, + "loss": 0.2183, + "step": 616 + }, + { + "epoch": 0.7299615498373262, + "grad_norm": 1.8791176080703735, + "learning_rate": 4.22875e-06, + "loss": 0.225, + "step": 617 + }, + { + "epoch": 0.7311446317657497, + "grad_norm": 1.7153689861297607, + "learning_rate": 4.227500000000001e-06, + "loss": 0.2066, + "step": 618 + }, + { + "epoch": 0.7323277136941734, + "grad_norm": 1.9367806911468506, + "learning_rate": 4.22625e-06, + "loss": 0.2348, + "step": 619 + }, + { + "epoch": 0.7335107956225969, + "grad_norm": 1.567488193511963, + "learning_rate": 4.225e-06, + "loss": 0.2067, + "step": 620 + }, + { + "epoch": 0.7346938775510204, + "grad_norm": 1.8361769914627075, + "learning_rate": 4.22375e-06, + "loss": 0.2129, + "step": 621 + }, + { + "epoch": 0.735876959479444, + "grad_norm": 1.9291653633117676, + "learning_rate": 4.222500000000001e-06, + "loss": 0.2016, + "step": 622 + }, + { + "epoch": 0.7370600414078675, + "grad_norm": 1.6848399639129639, + "learning_rate": 4.2212500000000005e-06, + "loss": 0.2138, + "step": 623 + }, + { + "epoch": 0.738243123336291, + "grad_norm": 1.4745463132858276, + "learning_rate": 4.22e-06, + "loss": 0.201, + "step": 624 + }, + { + "epoch": 0.7394262052647146, + "grad_norm": 1.583138108253479, + "learning_rate": 4.21875e-06, + "loss": 0.1859, + "step": 625 + }, + { + "epoch": 0.7406092871931381, + "grad_norm": 1.6470855474472046, + "learning_rate": 4.217500000000001e-06, + "loss": 0.1878, + "step": 626 + }, + { + "epoch": 0.7417923691215617, + "grad_norm": 1.8487787246704102, + "learning_rate": 4.2162500000000005e-06, + "loss": 0.2484, + "step": 627 + }, + { + "epoch": 0.7429754510499852, + "grad_norm": 1.441617488861084, + "learning_rate": 4.215e-06, + "loss": 0.1678, + "step": 628 + }, + { + "epoch": 0.7441585329784087, + "grad_norm": 1.8216331005096436, + "learning_rate": 4.21375e-06, + "loss": 0.2216, + "step": 629 + }, + { + "epoch": 0.7453416149068323, + "grad_norm": 1.678432583808899, + "learning_rate": 4.212500000000001e-06, + "loss": 0.1874, + "step": 630 + }, + { + "epoch": 0.7465246968352558, + "grad_norm": 1.7492058277130127, + "learning_rate": 4.2112500000000005e-06, + "loss": 0.2325, + "step": 631 + }, + { + "epoch": 0.7477077787636793, + "grad_norm": 1.4885244369506836, + "learning_rate": 4.21e-06, + "loss": 0.2095, + "step": 632 + }, + { + "epoch": 0.748890860692103, + "grad_norm": 1.533721685409546, + "learning_rate": 4.20875e-06, + "loss": 0.1895, + "step": 633 + }, + { + "epoch": 0.7500739426205265, + "grad_norm": 1.4698472023010254, + "learning_rate": 4.207500000000001e-06, + "loss": 0.219, + "step": 634 + }, + { + "epoch": 0.75125702454895, + "grad_norm": 1.669909119606018, + "learning_rate": 4.206250000000001e-06, + "loss": 0.2129, + "step": 635 + }, + { + "epoch": 0.7524401064773736, + "grad_norm": 1.5079299211502075, + "learning_rate": 4.205e-06, + "loss": 0.2251, + "step": 636 + }, + { + "epoch": 0.7536231884057971, + "grad_norm": 1.5760862827301025, + "learning_rate": 4.20375e-06, + "loss": 0.2113, + "step": 637 + }, + { + "epoch": 0.7548062703342207, + "grad_norm": 1.9492591619491577, + "learning_rate": 4.202500000000001e-06, + "loss": 0.2043, + "step": 638 + }, + { + "epoch": 0.7559893522626442, + "grad_norm": 1.6637297868728638, + "learning_rate": 4.201250000000001e-06, + "loss": 0.2176, + "step": 639 + }, + { + "epoch": 0.7571724341910677, + "grad_norm": 1.479033350944519, + "learning_rate": 4.2000000000000004e-06, + "loss": 0.2203, + "step": 640 + }, + { + "epoch": 0.7583555161194913, + "grad_norm": 1.7824300527572632, + "learning_rate": 4.19875e-06, + "loss": 0.2073, + "step": 641 + }, + { + "epoch": 0.7595385980479148, + "grad_norm": 1.539556860923767, + "learning_rate": 4.1975e-06, + "loss": 0.1954, + "step": 642 + }, + { + "epoch": 0.7607216799763383, + "grad_norm": 1.486075758934021, + "learning_rate": 4.196250000000001e-06, + "loss": 0.2225, + "step": 643 + }, + { + "epoch": 0.7619047619047619, + "grad_norm": 1.7244817018508911, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.2075, + "step": 644 + }, + { + "epoch": 0.7630878438331854, + "grad_norm": 1.5843439102172852, + "learning_rate": 4.19375e-06, + "loss": 0.1966, + "step": 645 + }, + { + "epoch": 0.764270925761609, + "grad_norm": 1.8097472190856934, + "learning_rate": 4.1925e-06, + "loss": 0.2108, + "step": 646 + }, + { + "epoch": 0.7654540076900326, + "grad_norm": 1.424957275390625, + "learning_rate": 4.191250000000001e-06, + "loss": 0.1978, + "step": 647 + }, + { + "epoch": 0.7666370896184561, + "grad_norm": 1.8132444620132446, + "learning_rate": 4.1900000000000005e-06, + "loss": 0.2157, + "step": 648 + }, + { + "epoch": 0.7678201715468796, + "grad_norm": 1.6925345659255981, + "learning_rate": 4.18875e-06, + "loss": 0.2231, + "step": 649 + }, + { + "epoch": 0.7690032534753032, + "grad_norm": 1.8036776781082153, + "learning_rate": 4.1875e-06, + "loss": 0.2247, + "step": 650 + }, + { + "epoch": 0.7701863354037267, + "grad_norm": 1.7948217391967773, + "learning_rate": 4.186250000000001e-06, + "loss": 0.2002, + "step": 651 + }, + { + "epoch": 0.7713694173321503, + "grad_norm": 1.881303071975708, + "learning_rate": 4.185000000000001e-06, + "loss": 0.2226, + "step": 652 + }, + { + "epoch": 0.7725524992605738, + "grad_norm": 1.5174137353897095, + "learning_rate": 4.18375e-06, + "loss": 0.2041, + "step": 653 + }, + { + "epoch": 0.7737355811889973, + "grad_norm": 1.4104382991790771, + "learning_rate": 4.1825e-06, + "loss": 0.1741, + "step": 654 + }, + { + "epoch": 0.7749186631174209, + "grad_norm": 1.7884827852249146, + "learning_rate": 4.181250000000001e-06, + "loss": 0.2027, + "step": 655 + }, + { + "epoch": 0.7761017450458444, + "grad_norm": 1.4523649215698242, + "learning_rate": 4.18e-06, + "loss": 0.1851, + "step": 656 + }, + { + "epoch": 0.7772848269742679, + "grad_norm": 1.5733247995376587, + "learning_rate": 4.1787500000000004e-06, + "loss": 0.1879, + "step": 657 + }, + { + "epoch": 0.7784679089026915, + "grad_norm": 1.7354931831359863, + "learning_rate": 4.1775e-06, + "loss": 0.1958, + "step": 658 + }, + { + "epoch": 0.7796509908311151, + "grad_norm": 1.7125475406646729, + "learning_rate": 4.176250000000001e-06, + "loss": 0.2085, + "step": 659 + }, + { + "epoch": 0.7808340727595386, + "grad_norm": 1.608055591583252, + "learning_rate": 4.175e-06, + "loss": 0.2049, + "step": 660 + }, + { + "epoch": 0.7820171546879622, + "grad_norm": 1.6849596500396729, + "learning_rate": 4.1737500000000005e-06, + "loss": 0.1949, + "step": 661 + }, + { + "epoch": 0.7832002366163857, + "grad_norm": 1.6813418865203857, + "learning_rate": 4.1725e-06, + "loss": 0.1981, + "step": 662 + }, + { + "epoch": 0.7843833185448092, + "grad_norm": 1.4315128326416016, + "learning_rate": 4.171250000000001e-06, + "loss": 0.2022, + "step": 663 + }, + { + "epoch": 0.7855664004732328, + "grad_norm": 1.6383264064788818, + "learning_rate": 4.17e-06, + "loss": 0.1948, + "step": 664 + }, + { + "epoch": 0.7867494824016563, + "grad_norm": 1.4421443939208984, + "learning_rate": 4.1687500000000005e-06, + "loss": 0.2055, + "step": 665 + }, + { + "epoch": 0.7879325643300799, + "grad_norm": 1.4936401844024658, + "learning_rate": 4.1675e-06, + "loss": 0.1807, + "step": 666 + }, + { + "epoch": 0.7891156462585034, + "grad_norm": 2.0444540977478027, + "learning_rate": 4.166250000000001e-06, + "loss": 0.2155, + "step": 667 + }, + { + "epoch": 0.7902987281869269, + "grad_norm": 1.5178685188293457, + "learning_rate": 4.165e-06, + "loss": 0.2113, + "step": 668 + }, + { + "epoch": 0.7914818101153505, + "grad_norm": 1.7585806846618652, + "learning_rate": 4.163750000000001e-06, + "loss": 0.1977, + "step": 669 + }, + { + "epoch": 0.792664892043774, + "grad_norm": 1.4937082529067993, + "learning_rate": 4.1625e-06, + "loss": 0.2038, + "step": 670 + }, + { + "epoch": 0.7938479739721975, + "grad_norm": 1.389409065246582, + "learning_rate": 4.16125e-06, + "loss": 0.1821, + "step": 671 + }, + { + "epoch": 0.7950310559006211, + "grad_norm": 1.4361777305603027, + "learning_rate": 4.16e-06, + "loss": 0.1776, + "step": 672 + }, + { + "epoch": 0.7962141378290447, + "grad_norm": 1.5811891555786133, + "learning_rate": 4.158750000000001e-06, + "loss": 0.1883, + "step": 673 + }, + { + "epoch": 0.7973972197574682, + "grad_norm": 1.7972438335418701, + "learning_rate": 4.1575000000000004e-06, + "loss": 0.2193, + "step": 674 + }, + { + "epoch": 0.7985803016858918, + "grad_norm": 1.4746594429016113, + "learning_rate": 4.15625e-06, + "loss": 0.1827, + "step": 675 + }, + { + "epoch": 0.7997633836143153, + "grad_norm": 1.5941245555877686, + "learning_rate": 4.155e-06, + "loss": 0.2132, + "step": 676 + }, + { + "epoch": 0.8009464655427389, + "grad_norm": 1.7128826379776, + "learning_rate": 4.153750000000001e-06, + "loss": 0.1996, + "step": 677 + }, + { + "epoch": 0.8021295474711624, + "grad_norm": 1.544480562210083, + "learning_rate": 4.1525000000000005e-06, + "loss": 0.1922, + "step": 678 + }, + { + "epoch": 0.8033126293995859, + "grad_norm": 1.5864170789718628, + "learning_rate": 4.15125e-06, + "loss": 0.1944, + "step": 679 + }, + { + "epoch": 0.8044957113280095, + "grad_norm": 1.763724684715271, + "learning_rate": 4.15e-06, + "loss": 0.1954, + "step": 680 + }, + { + "epoch": 0.805678793256433, + "grad_norm": 1.5672308206558228, + "learning_rate": 4.14875e-06, + "loss": 0.2095, + "step": 681 + }, + { + "epoch": 0.8068618751848565, + "grad_norm": 1.7577580213546753, + "learning_rate": 4.1475000000000005e-06, + "loss": 0.2118, + "step": 682 + }, + { + "epoch": 0.8080449571132801, + "grad_norm": 1.3207273483276367, + "learning_rate": 4.14625e-06, + "loss": 0.1904, + "step": 683 + }, + { + "epoch": 0.8092280390417036, + "grad_norm": 1.6204581260681152, + "learning_rate": 4.145e-06, + "loss": 0.1949, + "step": 684 + }, + { + "epoch": 0.8104111209701271, + "grad_norm": 1.3613961935043335, + "learning_rate": 4.14375e-06, + "loss": 0.2027, + "step": 685 + }, + { + "epoch": 0.8115942028985508, + "grad_norm": 1.6147979497909546, + "learning_rate": 4.1425000000000006e-06, + "loss": 0.2088, + "step": 686 + }, + { + "epoch": 0.8127772848269743, + "grad_norm": 1.6304869651794434, + "learning_rate": 4.14125e-06, + "loss": 0.1972, + "step": 687 + }, + { + "epoch": 0.8139603667553978, + "grad_norm": 1.4524974822998047, + "learning_rate": 4.14e-06, + "loss": 0.1961, + "step": 688 + }, + { + "epoch": 0.8151434486838214, + "grad_norm": 1.6035929918289185, + "learning_rate": 4.13875e-06, + "loss": 0.2269, + "step": 689 + }, + { + "epoch": 0.8163265306122449, + "grad_norm": 1.6616710424423218, + "learning_rate": 4.137500000000001e-06, + "loss": 0.2301, + "step": 690 + }, + { + "epoch": 0.8175096125406685, + "grad_norm": 1.5941667556762695, + "learning_rate": 4.13625e-06, + "loss": 0.1904, + "step": 691 + }, + { + "epoch": 0.818692694469092, + "grad_norm": 1.4404820203781128, + "learning_rate": 4.135e-06, + "loss": 0.1872, + "step": 692 + }, + { + "epoch": 0.8198757763975155, + "grad_norm": 1.55125892162323, + "learning_rate": 4.13375e-06, + "loss": 0.2122, + "step": 693 + }, + { + "epoch": 0.8210588583259391, + "grad_norm": 1.7499891519546509, + "learning_rate": 4.132500000000001e-06, + "loss": 0.2028, + "step": 694 + }, + { + "epoch": 0.8222419402543626, + "grad_norm": 1.6863594055175781, + "learning_rate": 4.1312500000000005e-06, + "loss": 0.1904, + "step": 695 + }, + { + "epoch": 0.8234250221827861, + "grad_norm": 1.5932685136795044, + "learning_rate": 4.13e-06, + "loss": 0.1992, + "step": 696 + }, + { + "epoch": 0.8246081041112097, + "grad_norm": 1.7375249862670898, + "learning_rate": 4.12875e-06, + "loss": 0.2059, + "step": 697 + }, + { + "epoch": 0.8257911860396332, + "grad_norm": 1.5555740594863892, + "learning_rate": 4.127500000000001e-06, + "loss": 0.2002, + "step": 698 + }, + { + "epoch": 0.8269742679680568, + "grad_norm": 1.218538761138916, + "learning_rate": 4.1262500000000005e-06, + "loss": 0.2075, + "step": 699 + }, + { + "epoch": 0.8281573498964804, + "grad_norm": 1.6516135931015015, + "learning_rate": 4.125e-06, + "loss": 0.2381, + "step": 700 + }, + { + "epoch": 0.8293404318249039, + "grad_norm": 1.4867770671844482, + "learning_rate": 4.12375e-06, + "loss": 0.1822, + "step": 701 + }, + { + "epoch": 0.8305235137533274, + "grad_norm": 1.4768725633621216, + "learning_rate": 4.122500000000001e-06, + "loss": 0.1853, + "step": 702 + }, + { + "epoch": 0.831706595681751, + "grad_norm": 1.6220333576202393, + "learning_rate": 4.1212500000000006e-06, + "loss": 0.2193, + "step": 703 + }, + { + "epoch": 0.8328896776101745, + "grad_norm": 1.565120816230774, + "learning_rate": 4.12e-06, + "loss": 0.1935, + "step": 704 + }, + { + "epoch": 0.834072759538598, + "grad_norm": 1.3800395727157593, + "learning_rate": 4.11875e-06, + "loss": 0.1775, + "step": 705 + }, + { + "epoch": 0.8352558414670216, + "grad_norm": 2.015958309173584, + "learning_rate": 4.1175e-06, + "loss": 0.1983, + "step": 706 + }, + { + "epoch": 0.8364389233954451, + "grad_norm": 1.642169713973999, + "learning_rate": 4.116250000000001e-06, + "loss": 0.2149, + "step": 707 + }, + { + "epoch": 0.8376220053238687, + "grad_norm": 1.6592271327972412, + "learning_rate": 4.115e-06, + "loss": 0.2066, + "step": 708 + }, + { + "epoch": 0.8388050872522922, + "grad_norm": 1.7273470163345337, + "learning_rate": 4.11375e-06, + "loss": 0.2136, + "step": 709 + }, + { + "epoch": 0.8399881691807157, + "grad_norm": 1.5744801759719849, + "learning_rate": 4.1125e-06, + "loss": 0.1928, + "step": 710 + }, + { + "epoch": 0.8411712511091393, + "grad_norm": 1.9545096158981323, + "learning_rate": 4.111250000000001e-06, + "loss": 0.223, + "step": 711 + }, + { + "epoch": 0.8423543330375628, + "grad_norm": 1.8394873142242432, + "learning_rate": 4.1100000000000005e-06, + "loss": 0.1943, + "step": 712 + }, + { + "epoch": 0.8435374149659864, + "grad_norm": 1.558488368988037, + "learning_rate": 4.10875e-06, + "loss": 0.1809, + "step": 713 + }, + { + "epoch": 0.84472049689441, + "grad_norm": 1.3312753438949585, + "learning_rate": 4.1075e-06, + "loss": 0.1885, + "step": 714 + }, + { + "epoch": 0.8459035788228335, + "grad_norm": 1.748753309249878, + "learning_rate": 4.106250000000001e-06, + "loss": 0.2048, + "step": 715 + }, + { + "epoch": 0.847086660751257, + "grad_norm": 1.6895931959152222, + "learning_rate": 4.1050000000000005e-06, + "loss": 0.2057, + "step": 716 + }, + { + "epoch": 0.8482697426796806, + "grad_norm": 1.4981443881988525, + "learning_rate": 4.10375e-06, + "loss": 0.1863, + "step": 717 + }, + { + "epoch": 0.8494528246081041, + "grad_norm": 1.7747372388839722, + "learning_rate": 4.1025e-06, + "loss": 0.2267, + "step": 718 + }, + { + "epoch": 0.8506359065365277, + "grad_norm": 1.5853220224380493, + "learning_rate": 4.101250000000001e-06, + "loss": 0.1974, + "step": 719 + }, + { + "epoch": 0.8518189884649512, + "grad_norm": 1.4564052820205688, + "learning_rate": 4.1e-06, + "loss": 0.212, + "step": 720 + }, + { + "epoch": 0.8530020703933747, + "grad_norm": 1.4939684867858887, + "learning_rate": 4.09875e-06, + "loss": 0.2095, + "step": 721 + }, + { + "epoch": 0.8541851523217983, + "grad_norm": 1.5663779973983765, + "learning_rate": 4.0975e-06, + "loss": 0.1899, + "step": 722 + }, + { + "epoch": 0.8553682342502218, + "grad_norm": 1.7670660018920898, + "learning_rate": 4.096250000000001e-06, + "loss": 0.1868, + "step": 723 + }, + { + "epoch": 0.8565513161786453, + "grad_norm": 1.839751124382019, + "learning_rate": 4.095e-06, + "loss": 0.2228, + "step": 724 + }, + { + "epoch": 0.8577343981070689, + "grad_norm": 1.6987298727035522, + "learning_rate": 4.09375e-06, + "loss": 0.2065, + "step": 725 + }, + { + "epoch": 0.8589174800354925, + "grad_norm": 1.6792759895324707, + "learning_rate": 4.0925e-06, + "loss": 0.2056, + "step": 726 + }, + { + "epoch": 0.860100561963916, + "grad_norm": 1.6812931299209595, + "learning_rate": 4.091250000000001e-06, + "loss": 0.2075, + "step": 727 + }, + { + "epoch": 0.8612836438923396, + "grad_norm": 1.6147911548614502, + "learning_rate": 4.09e-06, + "loss": 0.1907, + "step": 728 + }, + { + "epoch": 0.8624667258207631, + "grad_norm": 1.7735944986343384, + "learning_rate": 4.0887500000000005e-06, + "loss": 0.1827, + "step": 729 + }, + { + "epoch": 0.8636498077491866, + "grad_norm": 2.1575064659118652, + "learning_rate": 4.0875e-06, + "loss": 0.2158, + "step": 730 + }, + { + "epoch": 0.8648328896776102, + "grad_norm": 1.338484764099121, + "learning_rate": 4.086250000000001e-06, + "loss": 0.1986, + "step": 731 + }, + { + "epoch": 0.8660159716060337, + "grad_norm": 1.4411345720291138, + "learning_rate": 4.085e-06, + "loss": 0.2086, + "step": 732 + }, + { + "epoch": 0.8671990535344573, + "grad_norm": 1.4886226654052734, + "learning_rate": 4.0837500000000005e-06, + "loss": 0.2045, + "step": 733 + }, + { + "epoch": 0.8683821354628808, + "grad_norm": 1.5424730777740479, + "learning_rate": 4.0825e-06, + "loss": 0.1882, + "step": 734 + }, + { + "epoch": 0.8695652173913043, + "grad_norm": 1.7392501831054688, + "learning_rate": 4.08125e-06, + "loss": 0.1806, + "step": 735 + }, + { + "epoch": 0.8707482993197279, + "grad_norm": 1.6336326599121094, + "learning_rate": 4.08e-06, + "loss": 0.1991, + "step": 736 + }, + { + "epoch": 0.8719313812481514, + "grad_norm": 1.6803793907165527, + "learning_rate": 4.0787500000000005e-06, + "loss": 0.1934, + "step": 737 + }, + { + "epoch": 0.8731144631765749, + "grad_norm": 1.6031644344329834, + "learning_rate": 4.0775e-06, + "loss": 0.1943, + "step": 738 + }, + { + "epoch": 0.8742975451049986, + "grad_norm": 1.6850547790527344, + "learning_rate": 4.07625e-06, + "loss": 0.1846, + "step": 739 + }, + { + "epoch": 0.8754806270334221, + "grad_norm": 1.4961118698120117, + "learning_rate": 4.075e-06, + "loss": 0.1752, + "step": 740 + }, + { + "epoch": 0.8766637089618456, + "grad_norm": 2.055217981338501, + "learning_rate": 4.073750000000001e-06, + "loss": 0.2274, + "step": 741 + }, + { + "epoch": 0.8778467908902692, + "grad_norm": 1.5056447982788086, + "learning_rate": 4.0725e-06, + "loss": 0.2018, + "step": 742 + }, + { + "epoch": 0.8790298728186927, + "grad_norm": 1.660111427307129, + "learning_rate": 4.07125e-06, + "loss": 0.1942, + "step": 743 + }, + { + "epoch": 0.8802129547471162, + "grad_norm": 1.6051355600357056, + "learning_rate": 4.07e-06, + "loss": 0.2112, + "step": 744 + }, + { + "epoch": 0.8813960366755398, + "grad_norm": 1.380437970161438, + "learning_rate": 4.06875e-06, + "loss": 0.1912, + "step": 745 + }, + { + "epoch": 0.8825791186039633, + "grad_norm": 1.826065182685852, + "learning_rate": 4.0675000000000004e-06, + "loss": 0.2262, + "step": 746 + }, + { + "epoch": 0.8837622005323869, + "grad_norm": 1.4162591695785522, + "learning_rate": 4.06625e-06, + "loss": 0.1979, + "step": 747 + }, + { + "epoch": 0.8849452824608104, + "grad_norm": 1.5869451761245728, + "learning_rate": 4.065e-06, + "loss": 0.1967, + "step": 748 + }, + { + "epoch": 0.8861283643892339, + "grad_norm": 1.602133870124817, + "learning_rate": 4.06375e-06, + "loss": 0.2202, + "step": 749 + }, + { + "epoch": 0.8873114463176575, + "grad_norm": 1.5584497451782227, + "learning_rate": 4.0625000000000005e-06, + "loss": 0.1916, + "step": 750 + }, + { + "epoch": 0.888494528246081, + "grad_norm": 1.6562703847885132, + "learning_rate": 4.06125e-06, + "loss": 0.1863, + "step": 751 + }, + { + "epoch": 0.8896776101745045, + "grad_norm": 1.5322407484054565, + "learning_rate": 4.060000000000001e-06, + "loss": 0.1899, + "step": 752 + }, + { + "epoch": 0.8908606921029282, + "grad_norm": 1.7646961212158203, + "learning_rate": 4.05875e-06, + "loss": 0.1781, + "step": 753 + }, + { + "epoch": 0.8920437740313517, + "grad_norm": 1.5790330171585083, + "learning_rate": 4.0575000000000005e-06, + "loss": 0.2122, + "step": 754 + }, + { + "epoch": 0.8932268559597752, + "grad_norm": 1.2962757349014282, + "learning_rate": 4.05625e-06, + "loss": 0.1797, + "step": 755 + }, + { + "epoch": 0.8944099378881988, + "grad_norm": 1.7255396842956543, + "learning_rate": 4.055000000000001e-06, + "loss": 0.2303, + "step": 756 + }, + { + "epoch": 0.8955930198166223, + "grad_norm": 1.7830543518066406, + "learning_rate": 4.05375e-06, + "loss": 0.1862, + "step": 757 + }, + { + "epoch": 0.8967761017450459, + "grad_norm": 1.694628357887268, + "learning_rate": 4.052500000000001e-06, + "loss": 0.2125, + "step": 758 + }, + { + "epoch": 0.8979591836734694, + "grad_norm": 1.2690776586532593, + "learning_rate": 4.05125e-06, + "loss": 0.1875, + "step": 759 + }, + { + "epoch": 0.8991422656018929, + "grad_norm": 1.5824447870254517, + "learning_rate": 4.05e-06, + "loss": 0.2073, + "step": 760 + }, + { + "epoch": 0.9003253475303165, + "grad_norm": 1.5796151161193848, + "learning_rate": 4.04875e-06, + "loss": 0.1806, + "step": 761 + }, + { + "epoch": 0.90150842945874, + "grad_norm": 1.5095655918121338, + "learning_rate": 4.047500000000001e-06, + "loss": 0.2066, + "step": 762 + }, + { + "epoch": 0.9026915113871635, + "grad_norm": 1.6623122692108154, + "learning_rate": 4.0462500000000004e-06, + "loss": 0.1854, + "step": 763 + }, + { + "epoch": 0.9038745933155871, + "grad_norm": 1.4722758531570435, + "learning_rate": 4.045e-06, + "loss": 0.2156, + "step": 764 + }, + { + "epoch": 0.9050576752440106, + "grad_norm": 1.529905915260315, + "learning_rate": 4.04375e-06, + "loss": 0.1858, + "step": 765 + }, + { + "epoch": 0.9062407571724342, + "grad_norm": 1.6255297660827637, + "learning_rate": 4.042500000000001e-06, + "loss": 0.1881, + "step": 766 + }, + { + "epoch": 0.9074238391008578, + "grad_norm": 1.8210831880569458, + "learning_rate": 4.0412500000000005e-06, + "loss": 0.2277, + "step": 767 + }, + { + "epoch": 0.9086069210292813, + "grad_norm": 1.6574468612670898, + "learning_rate": 4.04e-06, + "loss": 0.2095, + "step": 768 + }, + { + "epoch": 0.9097900029577048, + "grad_norm": 1.8257920742034912, + "learning_rate": 4.03875e-06, + "loss": 0.1895, + "step": 769 + }, + { + "epoch": 0.9109730848861284, + "grad_norm": 1.456499457359314, + "learning_rate": 4.037500000000001e-06, + "loss": 0.1725, + "step": 770 + }, + { + "epoch": 0.9121561668145519, + "grad_norm": 1.5966291427612305, + "learning_rate": 4.0362500000000005e-06, + "loss": 0.1899, + "step": 771 + }, + { + "epoch": 0.9133392487429755, + "grad_norm": 1.8639682531356812, + "learning_rate": 4.035e-06, + "loss": 0.248, + "step": 772 + }, + { + "epoch": 0.914522330671399, + "grad_norm": 1.4585343599319458, + "learning_rate": 4.03375e-06, + "loss": 0.214, + "step": 773 + }, + { + "epoch": 0.9157054125998225, + "grad_norm": 1.4525967836380005, + "learning_rate": 4.0325e-06, + "loss": 0.2013, + "step": 774 + }, + { + "epoch": 0.9168884945282461, + "grad_norm": 1.4181418418884277, + "learning_rate": 4.031250000000001e-06, + "loss": 0.1856, + "step": 775 + }, + { + "epoch": 0.9180715764566696, + "grad_norm": 1.7247432470321655, + "learning_rate": 4.03e-06, + "loss": 0.1965, + "step": 776 + }, + { + "epoch": 0.9192546583850931, + "grad_norm": 1.6032253503799438, + "learning_rate": 4.02875e-06, + "loss": 0.2236, + "step": 777 + }, + { + "epoch": 0.9204377403135167, + "grad_norm": 1.4677108526229858, + "learning_rate": 4.0275e-06, + "loss": 0.2293, + "step": 778 + }, + { + "epoch": 0.9216208222419403, + "grad_norm": 1.8496289253234863, + "learning_rate": 4.026250000000001e-06, + "loss": 0.2013, + "step": 779 + }, + { + "epoch": 0.9228039041703638, + "grad_norm": 1.4107931852340698, + "learning_rate": 4.0250000000000004e-06, + "loss": 0.209, + "step": 780 + }, + { + "epoch": 0.9239869860987874, + "grad_norm": 1.5042765140533447, + "learning_rate": 4.02375e-06, + "loss": 0.2113, + "step": 781 + }, + { + "epoch": 0.9251700680272109, + "grad_norm": 1.797487497329712, + "learning_rate": 4.0225e-06, + "loss": 0.1858, + "step": 782 + }, + { + "epoch": 0.9263531499556344, + "grad_norm": 1.6710355281829834, + "learning_rate": 4.021250000000001e-06, + "loss": 0.1928, + "step": 783 + }, + { + "epoch": 0.927536231884058, + "grad_norm": 1.659069299697876, + "learning_rate": 4.0200000000000005e-06, + "loss": 0.2106, + "step": 784 + }, + { + "epoch": 0.9287193138124815, + "grad_norm": 1.6825233697891235, + "learning_rate": 4.01875e-06, + "loss": 0.206, + "step": 785 + }, + { + "epoch": 0.929902395740905, + "grad_norm": 1.4497143030166626, + "learning_rate": 4.0175e-06, + "loss": 0.1894, + "step": 786 + }, + { + "epoch": 0.9310854776693286, + "grad_norm": 1.5668892860412598, + "learning_rate": 4.016250000000001e-06, + "loss": 0.2142, + "step": 787 + }, + { + "epoch": 0.9322685595977521, + "grad_norm": 1.8686243295669556, + "learning_rate": 4.0150000000000005e-06, + "loss": 0.2139, + "step": 788 + }, + { + "epoch": 0.9334516415261757, + "grad_norm": 1.4304335117340088, + "learning_rate": 4.01375e-06, + "loss": 0.1809, + "step": 789 + }, + { + "epoch": 0.9346347234545992, + "grad_norm": 1.756691813468933, + "learning_rate": 4.0125e-06, + "loss": 0.1803, + "step": 790 + }, + { + "epoch": 0.9358178053830227, + "grad_norm": 1.9744210243225098, + "learning_rate": 4.011250000000001e-06, + "loss": 0.2234, + "step": 791 + }, + { + "epoch": 0.9370008873114463, + "grad_norm": 1.359911322593689, + "learning_rate": 4.0100000000000006e-06, + "loss": 0.19, + "step": 792 + }, + { + "epoch": 0.9381839692398699, + "grad_norm": 1.932350993156433, + "learning_rate": 4.00875e-06, + "loss": 0.2315, + "step": 793 + }, + { + "epoch": 0.9393670511682934, + "grad_norm": 1.4794167280197144, + "learning_rate": 4.0075e-06, + "loss": 0.1984, + "step": 794 + }, + { + "epoch": 0.940550133096717, + "grad_norm": 1.701920747756958, + "learning_rate": 4.006250000000001e-06, + "loss": 0.2182, + "step": 795 + }, + { + "epoch": 0.9417332150251405, + "grad_norm": 1.662898063659668, + "learning_rate": 4.005000000000001e-06, + "loss": 0.2117, + "step": 796 + }, + { + "epoch": 0.942916296953564, + "grad_norm": 2.2090184688568115, + "learning_rate": 4.00375e-06, + "loss": 0.2044, + "step": 797 + }, + { + "epoch": 0.9440993788819876, + "grad_norm": 1.6650887727737427, + "learning_rate": 4.0025e-06, + "loss": 0.2246, + "step": 798 + }, + { + "epoch": 0.9452824608104111, + "grad_norm": 1.6670759916305542, + "learning_rate": 4.00125e-06, + "loss": 0.2158, + "step": 799 + }, + { + "epoch": 0.9464655427388347, + "grad_norm": 1.8538198471069336, + "learning_rate": 4.000000000000001e-06, + "loss": 0.1933, + "step": 800 + }, + { + "epoch": 0.9476486246672582, + "grad_norm": 1.420890212059021, + "learning_rate": 3.9987500000000005e-06, + "loss": 0.2082, + "step": 801 + }, + { + "epoch": 0.9488317065956817, + "grad_norm": 1.548732042312622, + "learning_rate": 3.9975e-06, + "loss": 0.1875, + "step": 802 + }, + { + "epoch": 0.9500147885241053, + "grad_norm": 1.7208634614944458, + "learning_rate": 3.99625e-06, + "loss": 0.2155, + "step": 803 + }, + { + "epoch": 0.9511978704525288, + "grad_norm": 1.3889641761779785, + "learning_rate": 3.995000000000001e-06, + "loss": 0.1715, + "step": 804 + }, + { + "epoch": 0.9523809523809523, + "grad_norm": 1.6771676540374756, + "learning_rate": 3.9937500000000005e-06, + "loss": 0.1814, + "step": 805 + }, + { + "epoch": 0.953564034309376, + "grad_norm": 1.3889490365982056, + "learning_rate": 3.9925e-06, + "loss": 0.209, + "step": 806 + }, + { + "epoch": 0.9547471162377995, + "grad_norm": 1.5031368732452393, + "learning_rate": 3.99125e-06, + "loss": 0.223, + "step": 807 + }, + { + "epoch": 0.955930198166223, + "grad_norm": 1.4487813711166382, + "learning_rate": 3.990000000000001e-06, + "loss": 0.153, + "step": 808 + }, + { + "epoch": 0.9571132800946466, + "grad_norm": 1.5905200242996216, + "learning_rate": 3.98875e-06, + "loss": 0.1784, + "step": 809 + }, + { + "epoch": 0.9582963620230701, + "grad_norm": 1.3594461679458618, + "learning_rate": 3.9875e-06, + "loss": 0.1971, + "step": 810 + }, + { + "epoch": 0.9594794439514936, + "grad_norm": 1.6165438890457153, + "learning_rate": 3.98625e-06, + "loss": 0.1923, + "step": 811 + }, + { + "epoch": 0.9606625258799172, + "grad_norm": 1.558210849761963, + "learning_rate": 3.985000000000001e-06, + "loss": 0.2002, + "step": 812 + }, + { + "epoch": 0.9618456078083407, + "grad_norm": 1.7787020206451416, + "learning_rate": 3.98375e-06, + "loss": 0.1941, + "step": 813 + }, + { + "epoch": 0.9630286897367643, + "grad_norm": 1.6371811628341675, + "learning_rate": 3.9825e-06, + "loss": 0.2013, + "step": 814 + }, + { + "epoch": 0.9642117716651878, + "grad_norm": 1.4538682699203491, + "learning_rate": 3.98125e-06, + "loss": 0.1627, + "step": 815 + }, + { + "epoch": 0.9653948535936113, + "grad_norm": 1.7019946575164795, + "learning_rate": 3.980000000000001e-06, + "loss": 0.1825, + "step": 816 + }, + { + "epoch": 0.9665779355220349, + "grad_norm": 1.4935396909713745, + "learning_rate": 3.97875e-06, + "loss": 0.1932, + "step": 817 + }, + { + "epoch": 0.9677610174504584, + "grad_norm": 1.6102243661880493, + "learning_rate": 3.9775000000000005e-06, + "loss": 0.1734, + "step": 818 + }, + { + "epoch": 0.968944099378882, + "grad_norm": 1.5770198106765747, + "learning_rate": 3.97625e-06, + "loss": 0.2064, + "step": 819 + }, + { + "epoch": 0.9701271813073056, + "grad_norm": 1.8061292171478271, + "learning_rate": 3.975000000000001e-06, + "loss": 0.1889, + "step": 820 + }, + { + "epoch": 0.9713102632357291, + "grad_norm": 1.5744237899780273, + "learning_rate": 3.97375e-06, + "loss": 0.2057, + "step": 821 + }, + { + "epoch": 0.9724933451641526, + "grad_norm": 1.8933345079421997, + "learning_rate": 3.9725000000000005e-06, + "loss": 0.2289, + "step": 822 + }, + { + "epoch": 0.9736764270925762, + "grad_norm": 1.5591371059417725, + "learning_rate": 3.97125e-06, + "loss": 0.1959, + "step": 823 + }, + { + "epoch": 0.9748595090209997, + "grad_norm": 1.456443428993225, + "learning_rate": 3.97e-06, + "loss": 0.1989, + "step": 824 + }, + { + "epoch": 0.9760425909494232, + "grad_norm": 1.6241130828857422, + "learning_rate": 3.96875e-06, + "loss": 0.1906, + "step": 825 + }, + { + "epoch": 0.9772256728778468, + "grad_norm": 1.41323721408844, + "learning_rate": 3.9675000000000006e-06, + "loss": 0.1689, + "step": 826 + }, + { + "epoch": 0.9784087548062703, + "grad_norm": 1.6862794160842896, + "learning_rate": 3.96625e-06, + "loss": 0.1976, + "step": 827 + }, + { + "epoch": 0.9795918367346939, + "grad_norm": 1.5540764331817627, + "learning_rate": 3.965e-06, + "loss": 0.2013, + "step": 828 + }, + { + "epoch": 0.9807749186631174, + "grad_norm": 1.497370719909668, + "learning_rate": 3.96375e-06, + "loss": 0.2061, + "step": 829 + }, + { + "epoch": 0.9819580005915409, + "grad_norm": 1.4720622301101685, + "learning_rate": 3.962500000000001e-06, + "loss": 0.2072, + "step": 830 + }, + { + "epoch": 0.9831410825199645, + "grad_norm": 1.6613924503326416, + "learning_rate": 3.96125e-06, + "loss": 0.2014, + "step": 831 + }, + { + "epoch": 0.9843241644483881, + "grad_norm": 1.522898554801941, + "learning_rate": 3.96e-06, + "loss": 0.2081, + "step": 832 + }, + { + "epoch": 0.9855072463768116, + "grad_norm": 1.766837477684021, + "learning_rate": 3.95875e-06, + "loss": 0.2083, + "step": 833 + }, + { + "epoch": 0.9866903283052352, + "grad_norm": 1.3665306568145752, + "learning_rate": 3.957500000000001e-06, + "loss": 0.1896, + "step": 834 + }, + { + "epoch": 0.9878734102336587, + "grad_norm": 1.7022099494934082, + "learning_rate": 3.9562500000000004e-06, + "loss": 0.1924, + "step": 835 + }, + { + "epoch": 0.9890564921620822, + "grad_norm": 1.2004728317260742, + "learning_rate": 3.955e-06, + "loss": 0.1715, + "step": 836 + }, + { + "epoch": 0.9902395740905058, + "grad_norm": 1.6016632318496704, + "learning_rate": 3.95375e-06, + "loss": 0.1961, + "step": 837 + }, + { + "epoch": 0.9914226560189293, + "grad_norm": 1.5896190404891968, + "learning_rate": 3.9525e-06, + "loss": 0.1927, + "step": 838 + }, + { + "epoch": 0.9926057379473529, + "grad_norm": 1.560189127922058, + "learning_rate": 3.9512500000000005e-06, + "loss": 0.1986, + "step": 839 + }, + { + "epoch": 0.9937888198757764, + "grad_norm": 1.5898845195770264, + "learning_rate": 3.95e-06, + "loss": 0.2238, + "step": 840 + }, + { + "epoch": 0.9949719018041999, + "grad_norm": 1.959770679473877, + "learning_rate": 3.94875e-06, + "loss": 0.2215, + "step": 841 + }, + { + "epoch": 0.9961549837326235, + "grad_norm": 1.5834310054779053, + "learning_rate": 3.9475e-06, + "loss": 0.2062, + "step": 842 + }, + { + "epoch": 0.997338065661047, + "grad_norm": 1.5566670894622803, + "learning_rate": 3.9462500000000005e-06, + "loss": 0.1875, + "step": 843 + }, + { + "epoch": 0.9985211475894705, + "grad_norm": 1.610384464263916, + "learning_rate": 3.945e-06, + "loss": 0.1777, + "step": 844 + }, + { + "epoch": 0.9997042295178941, + "grad_norm": 1.5097066164016724, + "learning_rate": 3.94375e-06, + "loss": 0.1792, + "step": 845 + }, + { + "epoch": 1.0008873114463177, + "grad_norm": 1.4607676267623901, + "learning_rate": 3.9425e-06, + "loss": 0.1663, + "step": 846 + }, + { + "epoch": 1.0020703933747412, + "grad_norm": 1.5545316934585571, + "learning_rate": 3.941250000000001e-06, + "loss": 0.137, + "step": 847 + }, + { + "epoch": 1.0032534753031648, + "grad_norm": 1.5927177667617798, + "learning_rate": 3.94e-06, + "loss": 0.1414, + "step": 848 + }, + { + "epoch": 1.0044365572315883, + "grad_norm": 1.730399250984192, + "learning_rate": 3.93875e-06, + "loss": 0.1413, + "step": 849 + }, + { + "epoch": 1.0056196391600118, + "grad_norm": 1.5152299404144287, + "learning_rate": 3.9375e-06, + "loss": 0.1269, + "step": 850 + }, + { + "epoch": 1.0068027210884354, + "grad_norm": 1.5640554428100586, + "learning_rate": 3.936250000000001e-06, + "loss": 0.126, + "step": 851 + }, + { + "epoch": 1.007985803016859, + "grad_norm": 1.3630567789077759, + "learning_rate": 3.9350000000000004e-06, + "loss": 0.1255, + "step": 852 + }, + { + "epoch": 1.0091688849452825, + "grad_norm": 1.4786101579666138, + "learning_rate": 3.93375e-06, + "loss": 0.1269, + "step": 853 + }, + { + "epoch": 1.010351966873706, + "grad_norm": 1.5879299640655518, + "learning_rate": 3.9325e-06, + "loss": 0.1169, + "step": 854 + }, + { + "epoch": 1.0115350488021295, + "grad_norm": 1.4414206743240356, + "learning_rate": 3.931250000000001e-06, + "loss": 0.1112, + "step": 855 + }, + { + "epoch": 1.012718130730553, + "grad_norm": 1.540884256362915, + "learning_rate": 3.9300000000000005e-06, + "loss": 0.1325, + "step": 856 + }, + { + "epoch": 1.0139012126589766, + "grad_norm": 1.7065937519073486, + "learning_rate": 3.92875e-06, + "loss": 0.1456, + "step": 857 + }, + { + "epoch": 1.0150842945874001, + "grad_norm": 1.3394668102264404, + "learning_rate": 3.9275e-06, + "loss": 0.1265, + "step": 858 + }, + { + "epoch": 1.0162673765158237, + "grad_norm": 1.5018510818481445, + "learning_rate": 3.926250000000001e-06, + "loss": 0.1352, + "step": 859 + }, + { + "epoch": 1.0174504584442472, + "grad_norm": 1.3715797662734985, + "learning_rate": 3.9250000000000005e-06, + "loss": 0.1226, + "step": 860 + }, + { + "epoch": 1.0186335403726707, + "grad_norm": 1.3853869438171387, + "learning_rate": 3.92375e-06, + "loss": 0.1332, + "step": 861 + }, + { + "epoch": 1.0198166223010943, + "grad_norm": 1.4225016832351685, + "learning_rate": 3.9225e-06, + "loss": 0.1246, + "step": 862 + }, + { + "epoch": 1.0209997042295178, + "grad_norm": 1.8411242961883545, + "learning_rate": 3.92125e-06, + "loss": 0.1197, + "step": 863 + }, + { + "epoch": 1.0221827861579413, + "grad_norm": 1.7704321146011353, + "learning_rate": 3.920000000000001e-06, + "loss": 0.1277, + "step": 864 + }, + { + "epoch": 1.0233658680863649, + "grad_norm": 1.7261779308319092, + "learning_rate": 3.91875e-06, + "loss": 0.1398, + "step": 865 + }, + { + "epoch": 1.0245489500147886, + "grad_norm": 1.5654067993164062, + "learning_rate": 3.9175e-06, + "loss": 0.1242, + "step": 866 + }, + { + "epoch": 1.0257320319432122, + "grad_norm": 1.3489753007888794, + "learning_rate": 3.91625e-06, + "loss": 0.1126, + "step": 867 + }, + { + "epoch": 1.0269151138716357, + "grad_norm": 1.664570689201355, + "learning_rate": 3.915000000000001e-06, + "loss": 0.1254, + "step": 868 + }, + { + "epoch": 1.0280981958000592, + "grad_norm": 1.690885305404663, + "learning_rate": 3.9137500000000004e-06, + "loss": 0.127, + "step": 869 + }, + { + "epoch": 1.0292812777284828, + "grad_norm": 1.529438853263855, + "learning_rate": 3.9125e-06, + "loss": 0.1371, + "step": 870 + }, + { + "epoch": 1.0304643596569063, + "grad_norm": 1.598929524421692, + "learning_rate": 3.91125e-06, + "loss": 0.1249, + "step": 871 + }, + { + "epoch": 1.0316474415853298, + "grad_norm": 1.534091591835022, + "learning_rate": 3.910000000000001e-06, + "loss": 0.1191, + "step": 872 + }, + { + "epoch": 1.0328305235137534, + "grad_norm": 1.7261022329330444, + "learning_rate": 3.90875e-06, + "loss": 0.1309, + "step": 873 + }, + { + "epoch": 1.034013605442177, + "grad_norm": 1.5496859550476074, + "learning_rate": 3.9075e-06, + "loss": 0.1248, + "step": 874 + }, + { + "epoch": 1.0351966873706004, + "grad_norm": 1.592925786972046, + "learning_rate": 3.90625e-06, + "loss": 0.1199, + "step": 875 + }, + { + "epoch": 1.036379769299024, + "grad_norm": 1.5474638938903809, + "learning_rate": 3.905000000000001e-06, + "loss": 0.1306, + "step": 876 + }, + { + "epoch": 1.0375628512274475, + "grad_norm": 1.7254589796066284, + "learning_rate": 3.9037500000000005e-06, + "loss": 0.1309, + "step": 877 + }, + { + "epoch": 1.038745933155871, + "grad_norm": 1.5980417728424072, + "learning_rate": 3.9025e-06, + "loss": 0.1354, + "step": 878 + }, + { + "epoch": 1.0399290150842946, + "grad_norm": 1.4638845920562744, + "learning_rate": 3.90125e-06, + "loss": 0.1291, + "step": 879 + }, + { + "epoch": 1.0411120970127181, + "grad_norm": 1.5670627355575562, + "learning_rate": 3.900000000000001e-06, + "loss": 0.1261, + "step": 880 + }, + { + "epoch": 1.0422951789411417, + "grad_norm": 1.4395630359649658, + "learning_rate": 3.898750000000001e-06, + "loss": 0.1247, + "step": 881 + }, + { + "epoch": 1.0434782608695652, + "grad_norm": 1.6725918054580688, + "learning_rate": 3.8975e-06, + "loss": 0.1229, + "step": 882 + }, + { + "epoch": 1.0446613427979887, + "grad_norm": 1.8210841417312622, + "learning_rate": 3.89625e-06, + "loss": 0.1357, + "step": 883 + }, + { + "epoch": 1.0458444247264123, + "grad_norm": 1.572023630142212, + "learning_rate": 3.895000000000001e-06, + "loss": 0.1338, + "step": 884 + }, + { + "epoch": 1.0470275066548358, + "grad_norm": 1.719093918800354, + "learning_rate": 3.893750000000001e-06, + "loss": 0.1259, + "step": 885 + }, + { + "epoch": 1.0482105885832593, + "grad_norm": 1.3653970956802368, + "learning_rate": 3.8925000000000004e-06, + "loss": 0.1233, + "step": 886 + }, + { + "epoch": 1.0493936705116829, + "grad_norm": 1.528213381767273, + "learning_rate": 3.89125e-06, + "loss": 0.1333, + "step": 887 + }, + { + "epoch": 1.0505767524401064, + "grad_norm": 1.4623130559921265, + "learning_rate": 3.89e-06, + "loss": 0.1297, + "step": 888 + }, + { + "epoch": 1.05175983436853, + "grad_norm": 1.5873810052871704, + "learning_rate": 3.888750000000001e-06, + "loss": 0.1446, + "step": 889 + }, + { + "epoch": 1.0529429162969535, + "grad_norm": 1.7667055130004883, + "learning_rate": 3.8875000000000005e-06, + "loss": 0.1339, + "step": 890 + }, + { + "epoch": 1.054125998225377, + "grad_norm": 1.4597651958465576, + "learning_rate": 3.88625e-06, + "loss": 0.1228, + "step": 891 + }, + { + "epoch": 1.0553090801538008, + "grad_norm": 1.3840901851654053, + "learning_rate": 3.885e-06, + "loss": 0.1198, + "step": 892 + }, + { + "epoch": 1.0564921620822243, + "grad_norm": 1.595634937286377, + "learning_rate": 3.883750000000001e-06, + "loss": 0.136, + "step": 893 + }, + { + "epoch": 1.0576752440106478, + "grad_norm": 1.5623911619186401, + "learning_rate": 3.8825000000000005e-06, + "loss": 0.1188, + "step": 894 + }, + { + "epoch": 1.0588583259390714, + "grad_norm": 1.6958353519439697, + "learning_rate": 3.88125e-06, + "loss": 0.1376, + "step": 895 + }, + { + "epoch": 1.060041407867495, + "grad_norm": 1.5209115743637085, + "learning_rate": 3.88e-06, + "loss": 0.1259, + "step": 896 + }, + { + "epoch": 1.0612244897959184, + "grad_norm": 1.4908182621002197, + "learning_rate": 3.878750000000001e-06, + "loss": 0.1287, + "step": 897 + }, + { + "epoch": 1.062407571724342, + "grad_norm": 1.6207637786865234, + "learning_rate": 3.8775000000000006e-06, + "loss": 0.1275, + "step": 898 + }, + { + "epoch": 1.0635906536527655, + "grad_norm": 1.638773798942566, + "learning_rate": 3.87625e-06, + "loss": 0.1348, + "step": 899 + }, + { + "epoch": 1.064773735581189, + "grad_norm": 1.7876001596450806, + "learning_rate": 3.875e-06, + "loss": 0.1284, + "step": 900 + }, + { + "epoch": 1.0659568175096126, + "grad_norm": 1.439323902130127, + "learning_rate": 3.873750000000001e-06, + "loss": 0.1384, + "step": 901 + }, + { + "epoch": 1.067139899438036, + "grad_norm": 1.451987862586975, + "learning_rate": 3.8725e-06, + "loss": 0.1299, + "step": 902 + }, + { + "epoch": 1.0683229813664596, + "grad_norm": 1.3939586877822876, + "learning_rate": 3.87125e-06, + "loss": 0.1192, + "step": 903 + }, + { + "epoch": 1.0695060632948832, + "grad_norm": 1.5599257946014404, + "learning_rate": 3.87e-06, + "loss": 0.1289, + "step": 904 + }, + { + "epoch": 1.0706891452233067, + "grad_norm": 1.5900781154632568, + "learning_rate": 3.868750000000001e-06, + "loss": 0.1263, + "step": 905 + }, + { + "epoch": 1.0718722271517303, + "grad_norm": 1.9990553855895996, + "learning_rate": 3.8675e-06, + "loss": 0.1258, + "step": 906 + }, + { + "epoch": 1.0730553090801538, + "grad_norm": 1.4134851694107056, + "learning_rate": 3.8662500000000005e-06, + "loss": 0.1282, + "step": 907 + }, + { + "epoch": 1.0742383910085773, + "grad_norm": 1.5058811902999878, + "learning_rate": 3.865e-06, + "loss": 0.1302, + "step": 908 + }, + { + "epoch": 1.0754214729370009, + "grad_norm": 1.684141755104065, + "learning_rate": 3.863750000000001e-06, + "loss": 0.1278, + "step": 909 + }, + { + "epoch": 1.0766045548654244, + "grad_norm": 1.9097684621810913, + "learning_rate": 3.8625e-06, + "loss": 0.1245, + "step": 910 + }, + { + "epoch": 1.077787636793848, + "grad_norm": 1.5677382946014404, + "learning_rate": 3.8612500000000005e-06, + "loss": 0.1211, + "step": 911 + }, + { + "epoch": 1.0789707187222715, + "grad_norm": 1.4993963241577148, + "learning_rate": 3.86e-06, + "loss": 0.121, + "step": 912 + }, + { + "epoch": 1.080153800650695, + "grad_norm": 1.4101507663726807, + "learning_rate": 3.858750000000001e-06, + "loss": 0.1195, + "step": 913 + }, + { + "epoch": 1.0813368825791185, + "grad_norm": 1.552228331565857, + "learning_rate": 3.8575e-06, + "loss": 0.139, + "step": 914 + }, + { + "epoch": 1.082519964507542, + "grad_norm": 1.5412732362747192, + "learning_rate": 3.8562500000000006e-06, + "loss": 0.1325, + "step": 915 + }, + { + "epoch": 1.0837030464359656, + "grad_norm": 1.3636066913604736, + "learning_rate": 3.855e-06, + "loss": 0.1124, + "step": 916 + }, + { + "epoch": 1.0848861283643891, + "grad_norm": 1.4828418493270874, + "learning_rate": 3.85375e-06, + "loss": 0.1226, + "step": 917 + }, + { + "epoch": 1.0860692102928127, + "grad_norm": 1.5129079818725586, + "learning_rate": 3.8525e-06, + "loss": 0.1311, + "step": 918 + }, + { + "epoch": 1.0872522922212364, + "grad_norm": 1.60680091381073, + "learning_rate": 3.851250000000001e-06, + "loss": 0.1262, + "step": 919 + }, + { + "epoch": 1.08843537414966, + "grad_norm": 1.4580926895141602, + "learning_rate": 3.85e-06, + "loss": 0.1254, + "step": 920 + }, + { + "epoch": 1.0896184560780835, + "grad_norm": 1.6182721853256226, + "learning_rate": 3.84875e-06, + "loss": 0.1304, + "step": 921 + }, + { + "epoch": 1.090801538006507, + "grad_norm": 1.574212670326233, + "learning_rate": 3.8475e-06, + "loss": 0.1261, + "step": 922 + }, + { + "epoch": 1.0919846199349306, + "grad_norm": 1.5526423454284668, + "learning_rate": 3.846250000000001e-06, + "loss": 0.1202, + "step": 923 + }, + { + "epoch": 1.093167701863354, + "grad_norm": 2.1023406982421875, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.125, + "step": 924 + }, + { + "epoch": 1.0943507837917776, + "grad_norm": 1.4627243280410767, + "learning_rate": 3.84375e-06, + "loss": 0.1208, + "step": 925 + }, + { + "epoch": 1.0955338657202012, + "grad_norm": 1.554949164390564, + "learning_rate": 3.8425e-06, + "loss": 0.1282, + "step": 926 + }, + { + "epoch": 1.0967169476486247, + "grad_norm": 1.5215086936950684, + "learning_rate": 3.84125e-06, + "loss": 0.1296, + "step": 927 + }, + { + "epoch": 1.0979000295770482, + "grad_norm": 1.7638782262802124, + "learning_rate": 3.8400000000000005e-06, + "loss": 0.1396, + "step": 928 + }, + { + "epoch": 1.0990831115054718, + "grad_norm": 1.688349962234497, + "learning_rate": 3.83875e-06, + "loss": 0.1355, + "step": 929 + }, + { + "epoch": 1.1002661934338953, + "grad_norm": 1.2625313997268677, + "learning_rate": 3.8375e-06, + "loss": 0.1208, + "step": 930 + }, + { + "epoch": 1.1014492753623188, + "grad_norm": 1.6557972431182861, + "learning_rate": 3.83625e-06, + "loss": 0.131, + "step": 931 + }, + { + "epoch": 1.1026323572907424, + "grad_norm": 1.6518669128417969, + "learning_rate": 3.8350000000000006e-06, + "loss": 0.1392, + "step": 932 + }, + { + "epoch": 1.103815439219166, + "grad_norm": 1.7029277086257935, + "learning_rate": 3.83375e-06, + "loss": 0.1304, + "step": 933 + }, + { + "epoch": 1.1049985211475895, + "grad_norm": 1.4848511219024658, + "learning_rate": 3.8325e-06, + "loss": 0.1273, + "step": 934 + }, + { + "epoch": 1.106181603076013, + "grad_norm": 1.6099779605865479, + "learning_rate": 3.83125e-06, + "loss": 0.1253, + "step": 935 + }, + { + "epoch": 1.1073646850044365, + "grad_norm": 1.5807911157608032, + "learning_rate": 3.830000000000001e-06, + "loss": 0.1206, + "step": 936 + }, + { + "epoch": 1.10854776693286, + "grad_norm": 1.611435890197754, + "learning_rate": 3.82875e-06, + "loss": 0.1378, + "step": 937 + }, + { + "epoch": 1.1097308488612836, + "grad_norm": 1.5195038318634033, + "learning_rate": 3.8275e-06, + "loss": 0.131, + "step": 938 + }, + { + "epoch": 1.1109139307897071, + "grad_norm": 1.3802921772003174, + "learning_rate": 3.82625e-06, + "loss": 0.1218, + "step": 939 + }, + { + "epoch": 1.1120970127181307, + "grad_norm": 1.6278125047683716, + "learning_rate": 3.825000000000001e-06, + "loss": 0.1412, + "step": 940 + }, + { + "epoch": 1.1132800946465542, + "grad_norm": 1.5565327405929565, + "learning_rate": 3.8237500000000004e-06, + "loss": 0.1231, + "step": 941 + }, + { + "epoch": 1.1144631765749777, + "grad_norm": 1.8119821548461914, + "learning_rate": 3.8225e-06, + "loss": 0.1367, + "step": 942 + }, + { + "epoch": 1.1156462585034013, + "grad_norm": 1.739900827407837, + "learning_rate": 3.82125e-06, + "loss": 0.126, + "step": 943 + }, + { + "epoch": 1.1168293404318248, + "grad_norm": 1.4028265476226807, + "learning_rate": 3.820000000000001e-06, + "loss": 0.1276, + "step": 944 + }, + { + "epoch": 1.1180124223602483, + "grad_norm": 1.5332375764846802, + "learning_rate": 3.8187500000000005e-06, + "loss": 0.1225, + "step": 945 + }, + { + "epoch": 1.119195504288672, + "grad_norm": 1.5348633527755737, + "learning_rate": 3.8175e-06, + "loss": 0.1214, + "step": 946 + }, + { + "epoch": 1.1203785862170956, + "grad_norm": 1.5667204856872559, + "learning_rate": 3.81625e-06, + "loss": 0.1266, + "step": 947 + }, + { + "epoch": 1.1215616681455192, + "grad_norm": 1.4030245542526245, + "learning_rate": 3.815000000000001e-06, + "loss": 0.119, + "step": 948 + }, + { + "epoch": 1.1227447500739427, + "grad_norm": 1.3166753053665161, + "learning_rate": 3.8137500000000005e-06, + "loss": 0.1164, + "step": 949 + }, + { + "epoch": 1.1239278320023662, + "grad_norm": 1.5631295442581177, + "learning_rate": 3.8125e-06, + "loss": 0.1354, + "step": 950 + }, + { + "epoch": 1.1251109139307898, + "grad_norm": 1.6578530073165894, + "learning_rate": 3.81125e-06, + "loss": 0.1219, + "step": 951 + }, + { + "epoch": 1.1262939958592133, + "grad_norm": 1.7024977207183838, + "learning_rate": 3.8100000000000004e-06, + "loss": 0.1431, + "step": 952 + }, + { + "epoch": 1.1274770777876368, + "grad_norm": 1.5389904975891113, + "learning_rate": 3.8087500000000006e-06, + "loss": 0.1228, + "step": 953 + }, + { + "epoch": 1.1286601597160604, + "grad_norm": 1.888339877128601, + "learning_rate": 3.8075e-06, + "loss": 0.1301, + "step": 954 + }, + { + "epoch": 1.129843241644484, + "grad_norm": 1.5837042331695557, + "learning_rate": 3.80625e-06, + "loss": 0.1297, + "step": 955 + }, + { + "epoch": 1.1310263235729074, + "grad_norm": 1.4965708255767822, + "learning_rate": 3.8050000000000004e-06, + "loss": 0.1375, + "step": 956 + }, + { + "epoch": 1.132209405501331, + "grad_norm": 1.5138018131256104, + "learning_rate": 3.8037500000000006e-06, + "loss": 0.1211, + "step": 957 + }, + { + "epoch": 1.1333924874297545, + "grad_norm": 1.746869683265686, + "learning_rate": 3.8025e-06, + "loss": 0.1216, + "step": 958 + }, + { + "epoch": 1.134575569358178, + "grad_norm": 1.9778918027877808, + "learning_rate": 3.8012500000000002e-06, + "loss": 0.1316, + "step": 959 + }, + { + "epoch": 1.1357586512866016, + "grad_norm": 1.4542505741119385, + "learning_rate": 3.8000000000000005e-06, + "loss": 0.117, + "step": 960 + }, + { + "epoch": 1.1369417332150251, + "grad_norm": 1.6370503902435303, + "learning_rate": 3.7987500000000007e-06, + "loss": 0.1274, + "step": 961 + }, + { + "epoch": 1.1381248151434487, + "grad_norm": 1.7425072193145752, + "learning_rate": 3.7975e-06, + "loss": 0.1265, + "step": 962 + }, + { + "epoch": 1.1393078970718722, + "grad_norm": 1.541849136352539, + "learning_rate": 3.7962500000000003e-06, + "loss": 0.1406, + "step": 963 + }, + { + "epoch": 1.1404909790002957, + "grad_norm": 1.282357096672058, + "learning_rate": 3.7950000000000005e-06, + "loss": 0.1168, + "step": 964 + }, + { + "epoch": 1.1416740609287193, + "grad_norm": 1.4894678592681885, + "learning_rate": 3.7937500000000003e-06, + "loss": 0.1328, + "step": 965 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 1.7481800317764282, + "learning_rate": 3.7925e-06, + "loss": 0.1206, + "step": 966 + }, + { + "epoch": 1.1440402247855663, + "grad_norm": 1.5488405227661133, + "learning_rate": 3.7912500000000003e-06, + "loss": 0.1277, + "step": 967 + }, + { + "epoch": 1.1452233067139899, + "grad_norm": 1.5676923990249634, + "learning_rate": 3.79e-06, + "loss": 0.131, + "step": 968 + }, + { + "epoch": 1.1464063886424134, + "grad_norm": 1.8462034463882446, + "learning_rate": 3.7887500000000004e-06, + "loss": 0.1413, + "step": 969 + }, + { + "epoch": 1.147589470570837, + "grad_norm": 1.4963927268981934, + "learning_rate": 3.7875e-06, + "loss": 0.1277, + "step": 970 + }, + { + "epoch": 1.1487725524992607, + "grad_norm": 1.615854263305664, + "learning_rate": 3.7862500000000004e-06, + "loss": 0.1332, + "step": 971 + }, + { + "epoch": 1.149955634427684, + "grad_norm": 1.6165896654129028, + "learning_rate": 3.785e-06, + "loss": 0.1186, + "step": 972 + }, + { + "epoch": 1.1511387163561078, + "grad_norm": 1.7714965343475342, + "learning_rate": 3.7837500000000004e-06, + "loss": 0.133, + "step": 973 + }, + { + "epoch": 1.1523217982845313, + "grad_norm": 1.6568772792816162, + "learning_rate": 3.7825e-06, + "loss": 0.1282, + "step": 974 + }, + { + "epoch": 1.1535048802129548, + "grad_norm": 1.5904468297958374, + "learning_rate": 3.78125e-06, + "loss": 0.1309, + "step": 975 + }, + { + "epoch": 1.1546879621413784, + "grad_norm": 1.8618322610855103, + "learning_rate": 3.7800000000000002e-06, + "loss": 0.1427, + "step": 976 + }, + { + "epoch": 1.155871044069802, + "grad_norm": 1.7405513525009155, + "learning_rate": 3.7787500000000005e-06, + "loss": 0.1247, + "step": 977 + }, + { + "epoch": 1.1570541259982254, + "grad_norm": 1.7638050317764282, + "learning_rate": 3.7775000000000003e-06, + "loss": 0.1386, + "step": 978 + }, + { + "epoch": 1.158237207926649, + "grad_norm": 1.6512067317962646, + "learning_rate": 3.77625e-06, + "loss": 0.1245, + "step": 979 + }, + { + "epoch": 1.1594202898550725, + "grad_norm": 1.6244964599609375, + "learning_rate": 3.7750000000000003e-06, + "loss": 0.1208, + "step": 980 + }, + { + "epoch": 1.160603371783496, + "grad_norm": 1.2929716110229492, + "learning_rate": 3.7737500000000005e-06, + "loss": 0.1318, + "step": 981 + }, + { + "epoch": 1.1617864537119196, + "grad_norm": 1.7907743453979492, + "learning_rate": 3.7725e-06, + "loss": 0.122, + "step": 982 + }, + { + "epoch": 1.162969535640343, + "grad_norm": 1.565263271331787, + "learning_rate": 3.77125e-06, + "loss": 0.1198, + "step": 983 + }, + { + "epoch": 1.1641526175687666, + "grad_norm": 1.6273119449615479, + "learning_rate": 3.7700000000000003e-06, + "loss": 0.1291, + "step": 984 + }, + { + "epoch": 1.1653356994971902, + "grad_norm": 1.6684510707855225, + "learning_rate": 3.7687500000000006e-06, + "loss": 0.1306, + "step": 985 + }, + { + "epoch": 1.1665187814256137, + "grad_norm": 1.5227575302124023, + "learning_rate": 3.7675e-06, + "loss": 0.1212, + "step": 986 + }, + { + "epoch": 1.1677018633540373, + "grad_norm": 1.66986083984375, + "learning_rate": 3.76625e-06, + "loss": 0.123, + "step": 987 + }, + { + "epoch": 1.1688849452824608, + "grad_norm": 1.714493989944458, + "learning_rate": 3.7650000000000004e-06, + "loss": 0.1262, + "step": 988 + }, + { + "epoch": 1.1700680272108843, + "grad_norm": 1.488335371017456, + "learning_rate": 3.7637500000000006e-06, + "loss": 0.1148, + "step": 989 + }, + { + "epoch": 1.1712511091393079, + "grad_norm": 1.6159723997116089, + "learning_rate": 3.7625e-06, + "loss": 0.1295, + "step": 990 + }, + { + "epoch": 1.1724341910677314, + "grad_norm": 1.5783456563949585, + "learning_rate": 3.76125e-06, + "loss": 0.1233, + "step": 991 + }, + { + "epoch": 1.173617272996155, + "grad_norm": 1.5211812257766724, + "learning_rate": 3.7600000000000004e-06, + "loss": 0.1253, + "step": 992 + }, + { + "epoch": 1.1748003549245785, + "grad_norm": 1.9312084913253784, + "learning_rate": 3.7587500000000006e-06, + "loss": 0.1381, + "step": 993 + }, + { + "epoch": 1.175983436853002, + "grad_norm": 1.27653169631958, + "learning_rate": 3.7575e-06, + "loss": 0.1341, + "step": 994 + }, + { + "epoch": 1.1771665187814255, + "grad_norm": 1.3355271816253662, + "learning_rate": 3.7562500000000002e-06, + "loss": 0.1255, + "step": 995 + }, + { + "epoch": 1.178349600709849, + "grad_norm": 1.4467425346374512, + "learning_rate": 3.7550000000000005e-06, + "loss": 0.1302, + "step": 996 + }, + { + "epoch": 1.1795326826382726, + "grad_norm": 1.8044158220291138, + "learning_rate": 3.7537500000000003e-06, + "loss": 0.1414, + "step": 997 + }, + { + "epoch": 1.1807157645666964, + "grad_norm": 1.5622680187225342, + "learning_rate": 3.7525e-06, + "loss": 0.1234, + "step": 998 + }, + { + "epoch": 1.1818988464951197, + "grad_norm": 1.5954443216323853, + "learning_rate": 3.7512500000000003e-06, + "loss": 0.1178, + "step": 999 + }, + { + "epoch": 1.1830819284235434, + "grad_norm": 1.3906210660934448, + "learning_rate": 3.7500000000000005e-06, + "loss": 0.1246, + "step": 1000 + }, + { + "epoch": 1.184265010351967, + "grad_norm": 1.7755818367004395, + "learning_rate": 3.7487500000000003e-06, + "loss": 0.1334, + "step": 1001 + }, + { + "epoch": 1.1854480922803905, + "grad_norm": 1.5857216119766235, + "learning_rate": 3.7475000000000005e-06, + "loss": 0.1195, + "step": 1002 + }, + { + "epoch": 1.186631174208814, + "grad_norm": 1.6546953916549683, + "learning_rate": 3.7462500000000003e-06, + "loss": 0.1255, + "step": 1003 + }, + { + "epoch": 1.1878142561372376, + "grad_norm": 1.5238598585128784, + "learning_rate": 3.745e-06, + "loss": 0.1338, + "step": 1004 + }, + { + "epoch": 1.188997338065661, + "grad_norm": 1.5354591608047485, + "learning_rate": 3.7437500000000004e-06, + "loss": 0.1297, + "step": 1005 + }, + { + "epoch": 1.1901804199940846, + "grad_norm": 1.8391119241714478, + "learning_rate": 3.7425000000000006e-06, + "loss": 0.1361, + "step": 1006 + }, + { + "epoch": 1.1913635019225082, + "grad_norm": 1.5122896432876587, + "learning_rate": 3.74125e-06, + "loss": 0.1364, + "step": 1007 + }, + { + "epoch": 1.1925465838509317, + "grad_norm": 1.3886758089065552, + "learning_rate": 3.74e-06, + "loss": 0.1321, + "step": 1008 + }, + { + "epoch": 1.1937296657793552, + "grad_norm": 1.4924153089523315, + "learning_rate": 3.7387500000000004e-06, + "loss": 0.122, + "step": 1009 + }, + { + "epoch": 1.1949127477077788, + "grad_norm": 1.435950756072998, + "learning_rate": 3.7375000000000006e-06, + "loss": 0.1362, + "step": 1010 + }, + { + "epoch": 1.1960958296362023, + "grad_norm": 1.7854193449020386, + "learning_rate": 3.73625e-06, + "loss": 0.122, + "step": 1011 + }, + { + "epoch": 1.1972789115646258, + "grad_norm": 1.8019318580627441, + "learning_rate": 3.7350000000000002e-06, + "loss": 0.1264, + "step": 1012 + }, + { + "epoch": 1.1984619934930494, + "grad_norm": 1.4726303815841675, + "learning_rate": 3.7337500000000005e-06, + "loss": 0.1318, + "step": 1013 + }, + { + "epoch": 1.199645075421473, + "grad_norm": 1.6108293533325195, + "learning_rate": 3.7325000000000007e-06, + "loss": 0.1363, + "step": 1014 + }, + { + "epoch": 1.2008281573498965, + "grad_norm": 2.0222690105438232, + "learning_rate": 3.73125e-06, + "loss": 0.1368, + "step": 1015 + }, + { + "epoch": 1.20201123927832, + "grad_norm": 1.3891780376434326, + "learning_rate": 3.7300000000000003e-06, + "loss": 0.1344, + "step": 1016 + }, + { + "epoch": 1.2031943212067435, + "grad_norm": 1.6268746852874756, + "learning_rate": 3.7287500000000005e-06, + "loss": 0.1253, + "step": 1017 + }, + { + "epoch": 1.204377403135167, + "grad_norm": 1.5815043449401855, + "learning_rate": 3.7275000000000007e-06, + "loss": 0.1245, + "step": 1018 + }, + { + "epoch": 1.2055604850635906, + "grad_norm": 1.7551175355911255, + "learning_rate": 3.72625e-06, + "loss": 0.146, + "step": 1019 + }, + { + "epoch": 1.2067435669920141, + "grad_norm": 1.5023128986358643, + "learning_rate": 3.7250000000000003e-06, + "loss": 0.13, + "step": 1020 + }, + { + "epoch": 1.2079266489204377, + "grad_norm": 1.6033722162246704, + "learning_rate": 3.7237500000000006e-06, + "loss": 0.1326, + "step": 1021 + }, + { + "epoch": 1.2091097308488612, + "grad_norm": 1.4882923364639282, + "learning_rate": 3.7225000000000004e-06, + "loss": 0.1294, + "step": 1022 + }, + { + "epoch": 1.2102928127772847, + "grad_norm": 1.6634770631790161, + "learning_rate": 3.72125e-06, + "loss": 0.1228, + "step": 1023 + }, + { + "epoch": 1.2114758947057083, + "grad_norm": 1.5481011867523193, + "learning_rate": 3.7200000000000004e-06, + "loss": 0.1349, + "step": 1024 + }, + { + "epoch": 1.212658976634132, + "grad_norm": 1.3947616815567017, + "learning_rate": 3.7187500000000006e-06, + "loss": 0.1199, + "step": 1025 + }, + { + "epoch": 1.2138420585625553, + "grad_norm": 1.5992769002914429, + "learning_rate": 3.7175000000000004e-06, + "loss": 0.121, + "step": 1026 + }, + { + "epoch": 1.215025140490979, + "grad_norm": 1.683090329170227, + "learning_rate": 3.71625e-06, + "loss": 0.1275, + "step": 1027 + }, + { + "epoch": 1.2162082224194026, + "grad_norm": 1.7480064630508423, + "learning_rate": 3.7150000000000004e-06, + "loss": 0.1199, + "step": 1028 + }, + { + "epoch": 1.2173913043478262, + "grad_norm": 1.7127318382263184, + "learning_rate": 3.7137500000000002e-06, + "loss": 0.1207, + "step": 1029 + }, + { + "epoch": 1.2185743862762497, + "grad_norm": 1.9795665740966797, + "learning_rate": 3.7125000000000005e-06, + "loss": 0.1446, + "step": 1030 + }, + { + "epoch": 1.2197574682046732, + "grad_norm": 1.5555258989334106, + "learning_rate": 3.7112500000000003e-06, + "loss": 0.1212, + "step": 1031 + }, + { + "epoch": 1.2209405501330968, + "grad_norm": 1.6567680835723877, + "learning_rate": 3.7100000000000005e-06, + "loss": 0.1357, + "step": 1032 + }, + { + "epoch": 1.2221236320615203, + "grad_norm": 1.577582597732544, + "learning_rate": 3.7087500000000003e-06, + "loss": 0.1398, + "step": 1033 + }, + { + "epoch": 1.2233067139899438, + "grad_norm": 1.439808964729309, + "learning_rate": 3.7075000000000005e-06, + "loss": 0.1292, + "step": 1034 + }, + { + "epoch": 1.2244897959183674, + "grad_norm": 1.627306342124939, + "learning_rate": 3.7062500000000003e-06, + "loss": 0.125, + "step": 1035 + }, + { + "epoch": 1.225672877846791, + "grad_norm": 1.9445401430130005, + "learning_rate": 3.705e-06, + "loss": 0.1348, + "step": 1036 + }, + { + "epoch": 1.2268559597752144, + "grad_norm": 2.0726053714752197, + "learning_rate": 3.7037500000000003e-06, + "loss": 0.1244, + "step": 1037 + }, + { + "epoch": 1.228039041703638, + "grad_norm": 1.6830695867538452, + "learning_rate": 3.7025000000000005e-06, + "loss": 0.1339, + "step": 1038 + }, + { + "epoch": 1.2292221236320615, + "grad_norm": 1.8066555261611938, + "learning_rate": 3.70125e-06, + "loss": 0.1298, + "step": 1039 + }, + { + "epoch": 1.230405205560485, + "grad_norm": 1.658558964729309, + "learning_rate": 3.7e-06, + "loss": 0.1259, + "step": 1040 + }, + { + "epoch": 1.2315882874889086, + "grad_norm": 1.6936229467391968, + "learning_rate": 3.6987500000000004e-06, + "loss": 0.1298, + "step": 1041 + }, + { + "epoch": 1.2327713694173321, + "grad_norm": 1.6943538188934326, + "learning_rate": 3.6975000000000006e-06, + "loss": 0.144, + "step": 1042 + }, + { + "epoch": 1.2339544513457557, + "grad_norm": 2.1049861907958984, + "learning_rate": 3.69625e-06, + "loss": 0.1285, + "step": 1043 + }, + { + "epoch": 1.2351375332741792, + "grad_norm": 1.6036701202392578, + "learning_rate": 3.695e-06, + "loss": 0.1347, + "step": 1044 + }, + { + "epoch": 1.2363206152026027, + "grad_norm": 1.5913753509521484, + "learning_rate": 3.6937500000000004e-06, + "loss": 0.1251, + "step": 1045 + }, + { + "epoch": 1.2375036971310263, + "grad_norm": 1.6859318017959595, + "learning_rate": 3.6925000000000006e-06, + "loss": 0.122, + "step": 1046 + }, + { + "epoch": 1.2386867790594498, + "grad_norm": 1.4693753719329834, + "learning_rate": 3.69125e-06, + "loss": 0.1226, + "step": 1047 + }, + { + "epoch": 1.2398698609878733, + "grad_norm": 1.5733978748321533, + "learning_rate": 3.6900000000000002e-06, + "loss": 0.1372, + "step": 1048 + }, + { + "epoch": 1.2410529429162969, + "grad_norm": 2.0165650844573975, + "learning_rate": 3.6887500000000005e-06, + "loss": 0.132, + "step": 1049 + }, + { + "epoch": 1.2422360248447206, + "grad_norm": 1.5896422863006592, + "learning_rate": 3.6875000000000007e-06, + "loss": 0.1284, + "step": 1050 + }, + { + "epoch": 1.243419106773144, + "grad_norm": 1.6396112442016602, + "learning_rate": 3.68625e-06, + "loss": 0.1438, + "step": 1051 + }, + { + "epoch": 1.2446021887015677, + "grad_norm": 1.809408187866211, + "learning_rate": 3.6850000000000003e-06, + "loss": 0.131, + "step": 1052 + }, + { + "epoch": 1.245785270629991, + "grad_norm": 1.4872684478759766, + "learning_rate": 3.6837500000000005e-06, + "loss": 0.1355, + "step": 1053 + }, + { + "epoch": 1.2469683525584148, + "grad_norm": 1.6939865350723267, + "learning_rate": 3.6825000000000003e-06, + "loss": 0.1369, + "step": 1054 + }, + { + "epoch": 1.2481514344868383, + "grad_norm": 1.568205714225769, + "learning_rate": 3.68125e-06, + "loss": 0.1284, + "step": 1055 + }, + { + "epoch": 1.2493345164152618, + "grad_norm": 1.5654231309890747, + "learning_rate": 3.6800000000000003e-06, + "loss": 0.1359, + "step": 1056 + }, + { + "epoch": 1.2505175983436854, + "grad_norm": 1.4297429323196411, + "learning_rate": 3.6787500000000006e-06, + "loss": 0.1222, + "step": 1057 + }, + { + "epoch": 1.251700680272109, + "grad_norm": 1.7933324575424194, + "learning_rate": 3.6775000000000004e-06, + "loss": 0.132, + "step": 1058 + }, + { + "epoch": 1.2528837622005324, + "grad_norm": 1.838786244392395, + "learning_rate": 3.67625e-06, + "loss": 0.1296, + "step": 1059 + }, + { + "epoch": 1.254066844128956, + "grad_norm": 1.6346827745437622, + "learning_rate": 3.6750000000000004e-06, + "loss": 0.1282, + "step": 1060 + }, + { + "epoch": 1.2552499260573795, + "grad_norm": 1.6217823028564453, + "learning_rate": 3.67375e-06, + "loss": 0.1302, + "step": 1061 + }, + { + "epoch": 1.256433007985803, + "grad_norm": 1.532565951347351, + "learning_rate": 3.6725000000000004e-06, + "loss": 0.1302, + "step": 1062 + }, + { + "epoch": 1.2576160899142266, + "grad_norm": 1.8063253164291382, + "learning_rate": 3.6712500000000002e-06, + "loss": 0.1342, + "step": 1063 + }, + { + "epoch": 1.25879917184265, + "grad_norm": 1.8467254638671875, + "learning_rate": 3.6700000000000004e-06, + "loss": 0.1253, + "step": 1064 + }, + { + "epoch": 1.2599822537710736, + "grad_norm": 1.445910930633545, + "learning_rate": 3.6687500000000002e-06, + "loss": 0.1393, + "step": 1065 + }, + { + "epoch": 1.2611653356994972, + "grad_norm": 1.454780101776123, + "learning_rate": 3.6675000000000005e-06, + "loss": 0.1332, + "step": 1066 + }, + { + "epoch": 1.2623484176279207, + "grad_norm": 1.8868358135223389, + "learning_rate": 3.6662500000000003e-06, + "loss": 0.1394, + "step": 1067 + }, + { + "epoch": 1.2635314995563443, + "grad_norm": 1.5420104265213013, + "learning_rate": 3.665e-06, + "loss": 0.1124, + "step": 1068 + }, + { + "epoch": 1.2647145814847678, + "grad_norm": 1.4981539249420166, + "learning_rate": 3.6637500000000003e-06, + "loss": 0.1213, + "step": 1069 + }, + { + "epoch": 1.2658976634131913, + "grad_norm": 1.4259003400802612, + "learning_rate": 3.6625000000000005e-06, + "loss": 0.1191, + "step": 1070 + }, + { + "epoch": 1.2670807453416149, + "grad_norm": 1.8788647651672363, + "learning_rate": 3.66125e-06, + "loss": 0.1355, + "step": 1071 + }, + { + "epoch": 1.2682638272700384, + "grad_norm": 1.897626280784607, + "learning_rate": 3.66e-06, + "loss": 0.1301, + "step": 1072 + }, + { + "epoch": 1.269446909198462, + "grad_norm": 1.430627465248108, + "learning_rate": 3.6587500000000003e-06, + "loss": 0.1208, + "step": 1073 + }, + { + "epoch": 1.2706299911268855, + "grad_norm": 1.590336561203003, + "learning_rate": 3.6575000000000006e-06, + "loss": 0.1287, + "step": 1074 + }, + { + "epoch": 1.2718130730553092, + "grad_norm": 1.592786192893982, + "learning_rate": 3.65625e-06, + "loss": 0.1277, + "step": 1075 + }, + { + "epoch": 1.2729961549837325, + "grad_norm": 1.6330636739730835, + "learning_rate": 3.655e-06, + "loss": 0.1355, + "step": 1076 + }, + { + "epoch": 1.2741792369121563, + "grad_norm": 1.696073055267334, + "learning_rate": 3.6537500000000004e-06, + "loss": 0.1431, + "step": 1077 + }, + { + "epoch": 1.2753623188405796, + "grad_norm": 1.7294836044311523, + "learning_rate": 3.6525000000000006e-06, + "loss": 0.1348, + "step": 1078 + }, + { + "epoch": 1.2765454007690034, + "grad_norm": 1.7599605321884155, + "learning_rate": 3.65125e-06, + "loss": 0.1195, + "step": 1079 + }, + { + "epoch": 1.2777284826974267, + "grad_norm": 1.552577018737793, + "learning_rate": 3.65e-06, + "loss": 0.1292, + "step": 1080 + }, + { + "epoch": 1.2789115646258504, + "grad_norm": 1.6438440084457397, + "learning_rate": 3.6487500000000004e-06, + "loss": 0.1301, + "step": 1081 + }, + { + "epoch": 1.280094646554274, + "grad_norm": 1.6315746307373047, + "learning_rate": 3.6475000000000007e-06, + "loss": 0.1432, + "step": 1082 + }, + { + "epoch": 1.2812777284826975, + "grad_norm": 1.8404617309570312, + "learning_rate": 3.64625e-06, + "loss": 0.1306, + "step": 1083 + }, + { + "epoch": 1.282460810411121, + "grad_norm": 1.4895493984222412, + "learning_rate": 3.6450000000000003e-06, + "loss": 0.1168, + "step": 1084 + }, + { + "epoch": 1.2836438923395446, + "grad_norm": 1.5752488374710083, + "learning_rate": 3.6437500000000005e-06, + "loss": 0.1281, + "step": 1085 + }, + { + "epoch": 1.284826974267968, + "grad_norm": 1.756638526916504, + "learning_rate": 3.6425000000000003e-06, + "loss": 0.1344, + "step": 1086 + }, + { + "epoch": 1.2860100561963916, + "grad_norm": 1.3449690341949463, + "learning_rate": 3.64125e-06, + "loss": 0.1132, + "step": 1087 + }, + { + "epoch": 1.2871931381248152, + "grad_norm": 1.4565401077270508, + "learning_rate": 3.6400000000000003e-06, + "loss": 0.132, + "step": 1088 + }, + { + "epoch": 1.2883762200532387, + "grad_norm": 1.4078519344329834, + "learning_rate": 3.6387500000000005e-06, + "loss": 0.1201, + "step": 1089 + }, + { + "epoch": 1.2895593019816622, + "grad_norm": 1.570070743560791, + "learning_rate": 3.6375000000000003e-06, + "loss": 0.1317, + "step": 1090 + }, + { + "epoch": 1.2907423839100858, + "grad_norm": 1.6845135688781738, + "learning_rate": 3.63625e-06, + "loss": 0.1298, + "step": 1091 + }, + { + "epoch": 1.2919254658385093, + "grad_norm": 1.558335304260254, + "learning_rate": 3.6350000000000003e-06, + "loss": 0.118, + "step": 1092 + }, + { + "epoch": 1.2931085477669328, + "grad_norm": 1.6396114826202393, + "learning_rate": 3.63375e-06, + "loss": 0.1258, + "step": 1093 + }, + { + "epoch": 1.2942916296953564, + "grad_norm": 1.4986542463302612, + "learning_rate": 3.6325000000000004e-06, + "loss": 0.1397, + "step": 1094 + }, + { + "epoch": 1.29547471162378, + "grad_norm": 1.7439802885055542, + "learning_rate": 3.63125e-06, + "loss": 0.123, + "step": 1095 + }, + { + "epoch": 1.2966577935522035, + "grad_norm": 1.678720235824585, + "learning_rate": 3.6300000000000004e-06, + "loss": 0.1321, + "step": 1096 + }, + { + "epoch": 1.297840875480627, + "grad_norm": 1.652286171913147, + "learning_rate": 3.62875e-06, + "loss": 0.1334, + "step": 1097 + }, + { + "epoch": 1.2990239574090505, + "grad_norm": 1.6306040287017822, + "learning_rate": 3.6275000000000004e-06, + "loss": 0.1237, + "step": 1098 + }, + { + "epoch": 1.300207039337474, + "grad_norm": 2.253558874130249, + "learning_rate": 3.6262500000000002e-06, + "loss": 0.1379, + "step": 1099 + }, + { + "epoch": 1.3013901212658976, + "grad_norm": 1.4513882398605347, + "learning_rate": 3.625e-06, + "loss": 0.1352, + "step": 1100 + }, + { + "epoch": 1.3025732031943211, + "grad_norm": 1.502388596534729, + "learning_rate": 3.6237500000000002e-06, + "loss": 0.1422, + "step": 1101 + }, + { + "epoch": 1.3037562851227449, + "grad_norm": 1.491885781288147, + "learning_rate": 3.6225000000000005e-06, + "loss": 0.1276, + "step": 1102 + }, + { + "epoch": 1.3049393670511682, + "grad_norm": 1.7151025533676147, + "learning_rate": 3.62125e-06, + "loss": 0.1191, + "step": 1103 + }, + { + "epoch": 1.306122448979592, + "grad_norm": 1.5329397916793823, + "learning_rate": 3.62e-06, + "loss": 0.1389, + "step": 1104 + }, + { + "epoch": 1.3073055309080153, + "grad_norm": 1.661171793937683, + "learning_rate": 3.6187500000000003e-06, + "loss": 0.1268, + "step": 1105 + }, + { + "epoch": 1.308488612836439, + "grad_norm": 1.6615127325057983, + "learning_rate": 3.6175000000000005e-06, + "loss": 0.1388, + "step": 1106 + }, + { + "epoch": 1.3096716947648623, + "grad_norm": 1.903822898864746, + "learning_rate": 3.61625e-06, + "loss": 0.1273, + "step": 1107 + }, + { + "epoch": 1.310854776693286, + "grad_norm": 1.5279465913772583, + "learning_rate": 3.615e-06, + "loss": 0.1308, + "step": 1108 + }, + { + "epoch": 1.3120378586217096, + "grad_norm": 1.870363712310791, + "learning_rate": 3.6137500000000003e-06, + "loss": 0.1377, + "step": 1109 + }, + { + "epoch": 1.3132209405501332, + "grad_norm": 1.6977967023849487, + "learning_rate": 3.6125000000000006e-06, + "loss": 0.1126, + "step": 1110 + }, + { + "epoch": 1.3144040224785567, + "grad_norm": 1.5827949047088623, + "learning_rate": 3.61125e-06, + "loss": 0.1363, + "step": 1111 + }, + { + "epoch": 1.3155871044069802, + "grad_norm": 1.5780075788497925, + "learning_rate": 3.61e-06, + "loss": 0.1283, + "step": 1112 + }, + { + "epoch": 1.3167701863354038, + "grad_norm": 1.3944759368896484, + "learning_rate": 3.6087500000000004e-06, + "loss": 0.1171, + "step": 1113 + }, + { + "epoch": 1.3179532682638273, + "grad_norm": 1.5905603170394897, + "learning_rate": 3.6075000000000006e-06, + "loss": 0.126, + "step": 1114 + }, + { + "epoch": 1.3191363501922508, + "grad_norm": 1.7495099306106567, + "learning_rate": 3.60625e-06, + "loss": 0.137, + "step": 1115 + }, + { + "epoch": 1.3203194321206744, + "grad_norm": 1.6392492055892944, + "learning_rate": 3.6050000000000002e-06, + "loss": 0.1356, + "step": 1116 + }, + { + "epoch": 1.321502514049098, + "grad_norm": 1.8145921230316162, + "learning_rate": 3.6037500000000004e-06, + "loss": 0.1267, + "step": 1117 + }, + { + "epoch": 1.3226855959775214, + "grad_norm": 1.6999448537826538, + "learning_rate": 3.6025000000000002e-06, + "loss": 0.138, + "step": 1118 + }, + { + "epoch": 1.323868677905945, + "grad_norm": 1.2903354167938232, + "learning_rate": 3.60125e-06, + "loss": 0.1245, + "step": 1119 + }, + { + "epoch": 1.3250517598343685, + "grad_norm": 1.6310220956802368, + "learning_rate": 3.6000000000000003e-06, + "loss": 0.127, + "step": 1120 + }, + { + "epoch": 1.326234841762792, + "grad_norm": 1.6139092445373535, + "learning_rate": 3.5987500000000005e-06, + "loss": 0.1348, + "step": 1121 + }, + { + "epoch": 1.3274179236912156, + "grad_norm": 1.9158188104629517, + "learning_rate": 3.5975000000000003e-06, + "loss": 0.1278, + "step": 1122 + }, + { + "epoch": 1.3286010056196391, + "grad_norm": 1.4442896842956543, + "learning_rate": 3.59625e-06, + "loss": 0.12, + "step": 1123 + }, + { + "epoch": 1.3297840875480627, + "grad_norm": 1.541550874710083, + "learning_rate": 3.5950000000000003e-06, + "loss": 0.133, + "step": 1124 + }, + { + "epoch": 1.3309671694764862, + "grad_norm": 1.4513462781906128, + "learning_rate": 3.59375e-06, + "loss": 0.1254, + "step": 1125 + }, + { + "epoch": 1.3321502514049097, + "grad_norm": 1.4396653175354004, + "learning_rate": 3.5925000000000003e-06, + "loss": 0.1202, + "step": 1126 + }, + { + "epoch": 1.3333333333333333, + "grad_norm": 1.717483401298523, + "learning_rate": 3.5912500000000006e-06, + "loss": 0.1328, + "step": 1127 + }, + { + "epoch": 1.3345164152617568, + "grad_norm": 1.7318575382232666, + "learning_rate": 3.5900000000000004e-06, + "loss": 0.1321, + "step": 1128 + }, + { + "epoch": 1.3356994971901806, + "grad_norm": 1.4264169931411743, + "learning_rate": 3.58875e-06, + "loss": 0.1423, + "step": 1129 + }, + { + "epoch": 1.3368825791186039, + "grad_norm": 1.7376326322555542, + "learning_rate": 3.5875000000000004e-06, + "loss": 0.1305, + "step": 1130 + }, + { + "epoch": 1.3380656610470276, + "grad_norm": 1.5502262115478516, + "learning_rate": 3.5862500000000006e-06, + "loss": 0.1237, + "step": 1131 + }, + { + "epoch": 1.339248742975451, + "grad_norm": 1.3948478698730469, + "learning_rate": 3.585e-06, + "loss": 0.1208, + "step": 1132 + }, + { + "epoch": 1.3404318249038747, + "grad_norm": 1.643790364265442, + "learning_rate": 3.58375e-06, + "loss": 0.1238, + "step": 1133 + }, + { + "epoch": 1.341614906832298, + "grad_norm": 1.4631054401397705, + "learning_rate": 3.5825000000000004e-06, + "loss": 0.1346, + "step": 1134 + }, + { + "epoch": 1.3427979887607218, + "grad_norm": 1.5983502864837646, + "learning_rate": 3.5812500000000007e-06, + "loss": 0.1279, + "step": 1135 + }, + { + "epoch": 1.3439810706891453, + "grad_norm": 1.5456130504608154, + "learning_rate": 3.58e-06, + "loss": 0.1345, + "step": 1136 + }, + { + "epoch": 1.3451641526175688, + "grad_norm": 1.5223045349121094, + "learning_rate": 3.5787500000000003e-06, + "loss": 0.1302, + "step": 1137 + }, + { + "epoch": 1.3463472345459924, + "grad_norm": 1.52766752243042, + "learning_rate": 3.5775000000000005e-06, + "loss": 0.1255, + "step": 1138 + }, + { + "epoch": 1.347530316474416, + "grad_norm": 1.954994559288025, + "learning_rate": 3.5762500000000007e-06, + "loss": 0.1334, + "step": 1139 + }, + { + "epoch": 1.3487133984028394, + "grad_norm": 1.612317681312561, + "learning_rate": 3.575e-06, + "loss": 0.1299, + "step": 1140 + }, + { + "epoch": 1.349896480331263, + "grad_norm": 1.7186437845230103, + "learning_rate": 3.5737500000000003e-06, + "loss": 0.1281, + "step": 1141 + }, + { + "epoch": 1.3510795622596865, + "grad_norm": 1.4754208326339722, + "learning_rate": 3.5725000000000005e-06, + "loss": 0.1415, + "step": 1142 + }, + { + "epoch": 1.35226264418811, + "grad_norm": 1.5902824401855469, + "learning_rate": 3.5712500000000007e-06, + "loss": 0.1378, + "step": 1143 + }, + { + "epoch": 1.3534457261165336, + "grad_norm": 1.4900546073913574, + "learning_rate": 3.57e-06, + "loss": 0.1458, + "step": 1144 + }, + { + "epoch": 1.3546288080449571, + "grad_norm": 1.3753243684768677, + "learning_rate": 3.5687500000000003e-06, + "loss": 0.1185, + "step": 1145 + }, + { + "epoch": 1.3558118899733806, + "grad_norm": 1.292508602142334, + "learning_rate": 3.5675000000000006e-06, + "loss": 0.1269, + "step": 1146 + }, + { + "epoch": 1.3569949719018042, + "grad_norm": 1.6411457061767578, + "learning_rate": 3.5662500000000004e-06, + "loss": 0.1265, + "step": 1147 + }, + { + "epoch": 1.3581780538302277, + "grad_norm": 1.8571122884750366, + "learning_rate": 3.565e-06, + "loss": 0.1348, + "step": 1148 + }, + { + "epoch": 1.3593611357586513, + "grad_norm": 1.7001045942306519, + "learning_rate": 3.5637500000000004e-06, + "loss": 0.1217, + "step": 1149 + }, + { + "epoch": 1.3605442176870748, + "grad_norm": 1.4322460889816284, + "learning_rate": 3.5625e-06, + "loss": 0.1168, + "step": 1150 + }, + { + "epoch": 1.3617272996154983, + "grad_norm": 1.5621613264083862, + "learning_rate": 3.5612500000000004e-06, + "loss": 0.1367, + "step": 1151 + }, + { + "epoch": 1.3629103815439219, + "grad_norm": 1.70200777053833, + "learning_rate": 3.5600000000000002e-06, + "loss": 0.1312, + "step": 1152 + }, + { + "epoch": 1.3640934634723454, + "grad_norm": 1.646270751953125, + "learning_rate": 3.5587500000000004e-06, + "loss": 0.1259, + "step": 1153 + }, + { + "epoch": 1.365276545400769, + "grad_norm": 1.3117735385894775, + "learning_rate": 3.5575000000000002e-06, + "loss": 0.141, + "step": 1154 + }, + { + "epoch": 1.3664596273291925, + "grad_norm": 1.4881705045700073, + "learning_rate": 3.5562500000000005e-06, + "loss": 0.1177, + "step": 1155 + }, + { + "epoch": 1.3676427092576162, + "grad_norm": 1.6367751359939575, + "learning_rate": 3.5550000000000003e-06, + "loss": 0.1476, + "step": 1156 + }, + { + "epoch": 1.3688257911860395, + "grad_norm": 1.6853853464126587, + "learning_rate": 3.55375e-06, + "loss": 0.1273, + "step": 1157 + }, + { + "epoch": 1.3700088731144633, + "grad_norm": 1.9335606098175049, + "learning_rate": 3.5525000000000003e-06, + "loss": 0.1406, + "step": 1158 + }, + { + "epoch": 1.3711919550428866, + "grad_norm": 1.446752905845642, + "learning_rate": 3.5512500000000005e-06, + "loss": 0.1275, + "step": 1159 + }, + { + "epoch": 1.3723750369713104, + "grad_norm": 1.5854905843734741, + "learning_rate": 3.5500000000000003e-06, + "loss": 0.1317, + "step": 1160 + }, + { + "epoch": 1.3735581188997337, + "grad_norm": 1.6202325820922852, + "learning_rate": 3.54875e-06, + "loss": 0.1285, + "step": 1161 + }, + { + "epoch": 1.3747412008281574, + "grad_norm": 2.219072103500366, + "learning_rate": 3.5475000000000003e-06, + "loss": 0.1629, + "step": 1162 + }, + { + "epoch": 1.375924282756581, + "grad_norm": 1.6290174722671509, + "learning_rate": 3.5462500000000006e-06, + "loss": 0.1329, + "step": 1163 + }, + { + "epoch": 1.3771073646850045, + "grad_norm": 1.5328892469406128, + "learning_rate": 3.545e-06, + "loss": 0.1354, + "step": 1164 + }, + { + "epoch": 1.378290446613428, + "grad_norm": 1.488791584968567, + "learning_rate": 3.54375e-06, + "loss": 0.1338, + "step": 1165 + }, + { + "epoch": 1.3794735285418516, + "grad_norm": 1.5699701309204102, + "learning_rate": 3.5425000000000004e-06, + "loss": 0.1372, + "step": 1166 + }, + { + "epoch": 1.380656610470275, + "grad_norm": 1.4971731901168823, + "learning_rate": 3.5412500000000006e-06, + "loss": 0.1219, + "step": 1167 + }, + { + "epoch": 1.3818396923986986, + "grad_norm": 1.508593201637268, + "learning_rate": 3.54e-06, + "loss": 0.1238, + "step": 1168 + }, + { + "epoch": 1.3830227743271222, + "grad_norm": 1.4396504163742065, + "learning_rate": 3.5387500000000002e-06, + "loss": 0.1298, + "step": 1169 + }, + { + "epoch": 1.3842058562555457, + "grad_norm": 1.307847261428833, + "learning_rate": 3.5375000000000004e-06, + "loss": 0.1238, + "step": 1170 + }, + { + "epoch": 1.3853889381839692, + "grad_norm": 1.4901988506317139, + "learning_rate": 3.5362500000000007e-06, + "loss": 0.1267, + "step": 1171 + }, + { + "epoch": 1.3865720201123928, + "grad_norm": 1.8322440385818481, + "learning_rate": 3.535e-06, + "loss": 0.1297, + "step": 1172 + }, + { + "epoch": 1.3877551020408163, + "grad_norm": 1.6119638681411743, + "learning_rate": 3.5337500000000003e-06, + "loss": 0.1251, + "step": 1173 + }, + { + "epoch": 1.3889381839692398, + "grad_norm": 1.4196782112121582, + "learning_rate": 3.5325000000000005e-06, + "loss": 0.1213, + "step": 1174 + }, + { + "epoch": 1.3901212658976634, + "grad_norm": 1.53908109664917, + "learning_rate": 3.5312500000000007e-06, + "loss": 0.13, + "step": 1175 + }, + { + "epoch": 1.391304347826087, + "grad_norm": 1.776538610458374, + "learning_rate": 3.53e-06, + "loss": 0.1408, + "step": 1176 + }, + { + "epoch": 1.3924874297545105, + "grad_norm": 2.286358594894409, + "learning_rate": 3.5287500000000003e-06, + "loss": 0.139, + "step": 1177 + }, + { + "epoch": 1.393670511682934, + "grad_norm": 1.495193600654602, + "learning_rate": 3.5275000000000005e-06, + "loss": 0.1305, + "step": 1178 + }, + { + "epoch": 1.3948535936113575, + "grad_norm": 1.3600654602050781, + "learning_rate": 3.5262500000000003e-06, + "loss": 0.1238, + "step": 1179 + }, + { + "epoch": 1.396036675539781, + "grad_norm": 1.802796721458435, + "learning_rate": 3.525e-06, + "loss": 0.1379, + "step": 1180 + }, + { + "epoch": 1.3972197574682046, + "grad_norm": 1.590278148651123, + "learning_rate": 3.5237500000000004e-06, + "loss": 0.1302, + "step": 1181 + }, + { + "epoch": 1.3984028393966281, + "grad_norm": 1.7845808267593384, + "learning_rate": 3.5225e-06, + "loss": 0.1093, + "step": 1182 + }, + { + "epoch": 1.3995859213250519, + "grad_norm": 1.4317584037780762, + "learning_rate": 3.5212500000000004e-06, + "loss": 0.1308, + "step": 1183 + }, + { + "epoch": 1.4007690032534752, + "grad_norm": 1.5997761487960815, + "learning_rate": 3.52e-06, + "loss": 0.1373, + "step": 1184 + }, + { + "epoch": 1.401952085181899, + "grad_norm": 1.7386846542358398, + "learning_rate": 3.5187500000000004e-06, + "loss": 0.1328, + "step": 1185 + }, + { + "epoch": 1.4031351671103223, + "grad_norm": 1.4549872875213623, + "learning_rate": 3.5175e-06, + "loss": 0.1409, + "step": 1186 + }, + { + "epoch": 1.404318249038746, + "grad_norm": 1.899855613708496, + "learning_rate": 3.5162500000000004e-06, + "loss": 0.1265, + "step": 1187 + }, + { + "epoch": 1.4055013309671693, + "grad_norm": 1.385845422744751, + "learning_rate": 3.5150000000000002e-06, + "loss": 0.1258, + "step": 1188 + }, + { + "epoch": 1.406684412895593, + "grad_norm": 1.605231761932373, + "learning_rate": 3.51375e-06, + "loss": 0.1358, + "step": 1189 + }, + { + "epoch": 1.4078674948240166, + "grad_norm": 1.641336441040039, + "learning_rate": 3.5125000000000003e-06, + "loss": 0.1273, + "step": 1190 + }, + { + "epoch": 1.4090505767524402, + "grad_norm": 1.6675713062286377, + "learning_rate": 3.5112500000000005e-06, + "loss": 0.1389, + "step": 1191 + }, + { + "epoch": 1.4102336586808637, + "grad_norm": 2.0552802085876465, + "learning_rate": 3.5100000000000003e-06, + "loss": 0.128, + "step": 1192 + }, + { + "epoch": 1.4114167406092872, + "grad_norm": 1.6836698055267334, + "learning_rate": 3.50875e-06, + "loss": 0.1464, + "step": 1193 + }, + { + "epoch": 1.4125998225377108, + "grad_norm": 1.5497492551803589, + "learning_rate": 3.5075000000000003e-06, + "loss": 0.134, + "step": 1194 + }, + { + "epoch": 1.4137829044661343, + "grad_norm": 1.5456292629241943, + "learning_rate": 3.5062500000000005e-06, + "loss": 0.1261, + "step": 1195 + }, + { + "epoch": 1.4149659863945578, + "grad_norm": 1.52536141872406, + "learning_rate": 3.505e-06, + "loss": 0.1279, + "step": 1196 + }, + { + "epoch": 1.4161490683229814, + "grad_norm": 1.6498095989227295, + "learning_rate": 3.50375e-06, + "loss": 0.123, + "step": 1197 + }, + { + "epoch": 1.417332150251405, + "grad_norm": 1.4831663370132446, + "learning_rate": 3.5025000000000003e-06, + "loss": 0.1221, + "step": 1198 + }, + { + "epoch": 1.4185152321798284, + "grad_norm": 1.4253921508789062, + "learning_rate": 3.5012500000000006e-06, + "loss": 0.132, + "step": 1199 + }, + { + "epoch": 1.419698314108252, + "grad_norm": 1.566024661064148, + "learning_rate": 3.5e-06, + "loss": 0.1401, + "step": 1200 + }, + { + "epoch": 1.4208813960366755, + "grad_norm": 1.5801522731781006, + "learning_rate": 3.49875e-06, + "loss": 0.1278, + "step": 1201 + }, + { + "epoch": 1.422064477965099, + "grad_norm": 1.594814658164978, + "learning_rate": 3.4975000000000004e-06, + "loss": 0.1388, + "step": 1202 + }, + { + "epoch": 1.4232475598935226, + "grad_norm": 1.7192561626434326, + "learning_rate": 3.4962500000000006e-06, + "loss": 0.1508, + "step": 1203 + }, + { + "epoch": 1.4244306418219461, + "grad_norm": 1.8379285335540771, + "learning_rate": 3.495e-06, + "loss": 0.1276, + "step": 1204 + }, + { + "epoch": 1.4256137237503697, + "grad_norm": 1.87662672996521, + "learning_rate": 3.4937500000000002e-06, + "loss": 0.135, + "step": 1205 + }, + { + "epoch": 1.4267968056787932, + "grad_norm": 1.9310641288757324, + "learning_rate": 3.4925000000000004e-06, + "loss": 0.1287, + "step": 1206 + }, + { + "epoch": 1.4279798876072167, + "grad_norm": 1.720009684562683, + "learning_rate": 3.4912500000000007e-06, + "loss": 0.133, + "step": 1207 + }, + { + "epoch": 1.4291629695356405, + "grad_norm": 1.45742928981781, + "learning_rate": 3.49e-06, + "loss": 0.1344, + "step": 1208 + }, + { + "epoch": 1.4303460514640638, + "grad_norm": 1.4747129678726196, + "learning_rate": 3.4887500000000003e-06, + "loss": 0.12, + "step": 1209 + }, + { + "epoch": 1.4315291333924876, + "grad_norm": 1.5961334705352783, + "learning_rate": 3.4875000000000005e-06, + "loss": 0.1169, + "step": 1210 + }, + { + "epoch": 1.4327122153209109, + "grad_norm": 1.838158369064331, + "learning_rate": 3.4862500000000003e-06, + "loss": 0.139, + "step": 1211 + }, + { + "epoch": 1.4338952972493346, + "grad_norm": 1.6690236330032349, + "learning_rate": 3.485e-06, + "loss": 0.1258, + "step": 1212 + }, + { + "epoch": 1.435078379177758, + "grad_norm": 1.6403840780258179, + "learning_rate": 3.4837500000000003e-06, + "loss": 0.1324, + "step": 1213 + }, + { + "epoch": 1.4362614611061817, + "grad_norm": 1.929845929145813, + "learning_rate": 3.4825e-06, + "loss": 0.1319, + "step": 1214 + }, + { + "epoch": 1.4374445430346052, + "grad_norm": 1.4625489711761475, + "learning_rate": 3.4812500000000003e-06, + "loss": 0.1295, + "step": 1215 + }, + { + "epoch": 1.4386276249630288, + "grad_norm": 1.4935985803604126, + "learning_rate": 3.48e-06, + "loss": 0.1262, + "step": 1216 + }, + { + "epoch": 1.4398107068914523, + "grad_norm": 1.4362653493881226, + "learning_rate": 3.4787500000000004e-06, + "loss": 0.139, + "step": 1217 + }, + { + "epoch": 1.4409937888198758, + "grad_norm": 1.6384072303771973, + "learning_rate": 3.4775e-06, + "loss": 0.1394, + "step": 1218 + }, + { + "epoch": 1.4421768707482994, + "grad_norm": 1.8191174268722534, + "learning_rate": 3.4762500000000004e-06, + "loss": 0.1283, + "step": 1219 + }, + { + "epoch": 1.443359952676723, + "grad_norm": 1.4115757942199707, + "learning_rate": 3.475e-06, + "loss": 0.1336, + "step": 1220 + }, + { + "epoch": 1.4445430346051464, + "grad_norm": 1.3643068075180054, + "learning_rate": 3.47375e-06, + "loss": 0.1323, + "step": 1221 + }, + { + "epoch": 1.44572611653357, + "grad_norm": 1.4798321723937988, + "learning_rate": 3.4725e-06, + "loss": 0.1217, + "step": 1222 + }, + { + "epoch": 1.4469091984619935, + "grad_norm": 1.4897466897964478, + "learning_rate": 3.4712500000000004e-06, + "loss": 0.1238, + "step": 1223 + }, + { + "epoch": 1.448092280390417, + "grad_norm": 1.9903239011764526, + "learning_rate": 3.4700000000000002e-06, + "loss": 0.1362, + "step": 1224 + }, + { + "epoch": 1.4492753623188406, + "grad_norm": 1.400438666343689, + "learning_rate": 3.46875e-06, + "loss": 0.121, + "step": 1225 + }, + { + "epoch": 1.4504584442472641, + "grad_norm": 1.6418116092681885, + "learning_rate": 3.4675000000000003e-06, + "loss": 0.1221, + "step": 1226 + }, + { + "epoch": 1.4516415261756876, + "grad_norm": 1.356575846672058, + "learning_rate": 3.4662500000000005e-06, + "loss": 0.1174, + "step": 1227 + }, + { + "epoch": 1.4528246081041112, + "grad_norm": 1.5767966508865356, + "learning_rate": 3.465e-06, + "loss": 0.1365, + "step": 1228 + }, + { + "epoch": 1.4540076900325347, + "grad_norm": 1.4591554403305054, + "learning_rate": 3.46375e-06, + "loss": 0.1275, + "step": 1229 + }, + { + "epoch": 1.4551907719609583, + "grad_norm": 1.869500756263733, + "learning_rate": 3.4625000000000003e-06, + "loss": 0.1378, + "step": 1230 + }, + { + "epoch": 1.4563738538893818, + "grad_norm": 1.6662275791168213, + "learning_rate": 3.4612500000000005e-06, + "loss": 0.1318, + "step": 1231 + }, + { + "epoch": 1.4575569358178053, + "grad_norm": 1.6397216320037842, + "learning_rate": 3.46e-06, + "loss": 0.1305, + "step": 1232 + }, + { + "epoch": 1.4587400177462289, + "grad_norm": 1.3928391933441162, + "learning_rate": 3.45875e-06, + "loss": 0.1266, + "step": 1233 + }, + { + "epoch": 1.4599230996746524, + "grad_norm": 1.5611717700958252, + "learning_rate": 3.4575000000000004e-06, + "loss": 0.1363, + "step": 1234 + }, + { + "epoch": 1.4611061816030761, + "grad_norm": 1.6412303447723389, + "learning_rate": 3.4562500000000006e-06, + "loss": 0.1236, + "step": 1235 + }, + { + "epoch": 1.4622892635314995, + "grad_norm": 1.6755750179290771, + "learning_rate": 3.455e-06, + "loss": 0.1226, + "step": 1236 + }, + { + "epoch": 1.4634723454599232, + "grad_norm": 1.7662516832351685, + "learning_rate": 3.45375e-06, + "loss": 0.145, + "step": 1237 + }, + { + "epoch": 1.4646554273883465, + "grad_norm": 1.711784839630127, + "learning_rate": 3.4525000000000004e-06, + "loss": 0.1414, + "step": 1238 + }, + { + "epoch": 1.4658385093167703, + "grad_norm": 1.6295006275177002, + "learning_rate": 3.4512500000000006e-06, + "loss": 0.1186, + "step": 1239 + }, + { + "epoch": 1.4670215912451936, + "grad_norm": 1.408040165901184, + "learning_rate": 3.45e-06, + "loss": 0.1268, + "step": 1240 + }, + { + "epoch": 1.4682046731736174, + "grad_norm": 1.6011841297149658, + "learning_rate": 3.4487500000000002e-06, + "loss": 0.1231, + "step": 1241 + }, + { + "epoch": 1.469387755102041, + "grad_norm": 1.9084091186523438, + "learning_rate": 3.4475000000000005e-06, + "loss": 0.1433, + "step": 1242 + }, + { + "epoch": 1.4705708370304644, + "grad_norm": 1.5943597555160522, + "learning_rate": 3.4462500000000003e-06, + "loss": 0.1296, + "step": 1243 + }, + { + "epoch": 1.471753918958888, + "grad_norm": 1.330012321472168, + "learning_rate": 3.445e-06, + "loss": 0.1183, + "step": 1244 + }, + { + "epoch": 1.4729370008873115, + "grad_norm": 1.5377483367919922, + "learning_rate": 3.4437500000000003e-06, + "loss": 0.1265, + "step": 1245 + }, + { + "epoch": 1.474120082815735, + "grad_norm": 1.4315767288208008, + "learning_rate": 3.4425e-06, + "loss": 0.1214, + "step": 1246 + }, + { + "epoch": 1.4753031647441586, + "grad_norm": 1.5559818744659424, + "learning_rate": 3.4412500000000003e-06, + "loss": 0.1255, + "step": 1247 + }, + { + "epoch": 1.476486246672582, + "grad_norm": 1.5628515481948853, + "learning_rate": 3.44e-06, + "loss": 0.1264, + "step": 1248 + }, + { + "epoch": 1.4776693286010056, + "grad_norm": 1.447611927986145, + "learning_rate": 3.4387500000000003e-06, + "loss": 0.1317, + "step": 1249 + }, + { + "epoch": 1.4788524105294292, + "grad_norm": 1.736879825592041, + "learning_rate": 3.4375e-06, + "loss": 0.1283, + "step": 1250 + }, + { + "epoch": 1.4800354924578527, + "grad_norm": 1.7619876861572266, + "learning_rate": 3.4362500000000003e-06, + "loss": 0.1247, + "step": 1251 + }, + { + "epoch": 1.4812185743862762, + "grad_norm": 1.6373612880706787, + "learning_rate": 3.4350000000000006e-06, + "loss": 0.1254, + "step": 1252 + }, + { + "epoch": 1.4824016563146998, + "grad_norm": 1.565795660018921, + "learning_rate": 3.43375e-06, + "loss": 0.1129, + "step": 1253 + }, + { + "epoch": 1.4835847382431233, + "grad_norm": 1.5357458591461182, + "learning_rate": 3.4325e-06, + "loss": 0.1387, + "step": 1254 + }, + { + "epoch": 1.4847678201715468, + "grad_norm": 1.9255259037017822, + "learning_rate": 3.4312500000000004e-06, + "loss": 0.1415, + "step": 1255 + }, + { + "epoch": 1.4859509020999704, + "grad_norm": 1.4566707611083984, + "learning_rate": 3.4300000000000006e-06, + "loss": 0.1312, + "step": 1256 + }, + { + "epoch": 1.487133984028394, + "grad_norm": 2.0529234409332275, + "learning_rate": 3.42875e-06, + "loss": 0.1331, + "step": 1257 + }, + { + "epoch": 1.4883170659568175, + "grad_norm": 1.4404765367507935, + "learning_rate": 3.4275000000000002e-06, + "loss": 0.1351, + "step": 1258 + }, + { + "epoch": 1.489500147885241, + "grad_norm": 1.5901224613189697, + "learning_rate": 3.4262500000000004e-06, + "loss": 0.1209, + "step": 1259 + }, + { + "epoch": 1.4906832298136645, + "grad_norm": 1.4635899066925049, + "learning_rate": 3.4250000000000007e-06, + "loss": 0.1248, + "step": 1260 + }, + { + "epoch": 1.491866311742088, + "grad_norm": 1.3403890132904053, + "learning_rate": 3.42375e-06, + "loss": 0.1312, + "step": 1261 + }, + { + "epoch": 1.4930493936705118, + "grad_norm": 1.5730700492858887, + "learning_rate": 3.4225000000000003e-06, + "loss": 0.1321, + "step": 1262 + }, + { + "epoch": 1.4942324755989351, + "grad_norm": 1.40403413772583, + "learning_rate": 3.4212500000000005e-06, + "loss": 0.1326, + "step": 1263 + }, + { + "epoch": 1.4954155575273589, + "grad_norm": 1.6415314674377441, + "learning_rate": 3.4200000000000007e-06, + "loss": 0.1392, + "step": 1264 + }, + { + "epoch": 1.4965986394557822, + "grad_norm": 1.5355576276779175, + "learning_rate": 3.41875e-06, + "loss": 0.1404, + "step": 1265 + }, + { + "epoch": 1.497781721384206, + "grad_norm": 1.5298172235488892, + "learning_rate": 3.4175000000000003e-06, + "loss": 0.137, + "step": 1266 + }, + { + "epoch": 1.4989648033126293, + "grad_norm": 1.5886250734329224, + "learning_rate": 3.4162500000000005e-06, + "loss": 0.1249, + "step": 1267 + }, + { + "epoch": 1.500147885241053, + "grad_norm": 1.689842939376831, + "learning_rate": 3.4150000000000003e-06, + "loss": 0.1453, + "step": 1268 + }, + { + "epoch": 1.5013309671694763, + "grad_norm": 1.4077543020248413, + "learning_rate": 3.41375e-06, + "loss": 0.1281, + "step": 1269 + }, + { + "epoch": 1.5025140490979, + "grad_norm": 1.3884670734405518, + "learning_rate": 3.4125000000000004e-06, + "loss": 0.1338, + "step": 1270 + }, + { + "epoch": 1.5036971310263234, + "grad_norm": 1.5794848203659058, + "learning_rate": 3.4112500000000006e-06, + "loss": 0.1333, + "step": 1271 + }, + { + "epoch": 1.5048802129547472, + "grad_norm": 1.9224528074264526, + "learning_rate": 3.4100000000000004e-06, + "loss": 0.1239, + "step": 1272 + }, + { + "epoch": 1.5060632948831707, + "grad_norm": 1.5073069334030151, + "learning_rate": 3.40875e-06, + "loss": 0.1258, + "step": 1273 + }, + { + "epoch": 1.5072463768115942, + "grad_norm": 1.3602319955825806, + "learning_rate": 3.4075000000000004e-06, + "loss": 0.1227, + "step": 1274 + }, + { + "epoch": 1.5084294587400178, + "grad_norm": 1.381373643875122, + "learning_rate": 3.40625e-06, + "loss": 0.1201, + "step": 1275 + }, + { + "epoch": 1.5096125406684413, + "grad_norm": 1.6320122480392456, + "learning_rate": 3.4050000000000004e-06, + "loss": 0.126, + "step": 1276 + }, + { + "epoch": 1.5107956225968648, + "grad_norm": 1.4455208778381348, + "learning_rate": 3.4037500000000002e-06, + "loss": 0.1274, + "step": 1277 + }, + { + "epoch": 1.5119787045252884, + "grad_norm": 1.8690789937973022, + "learning_rate": 3.4025000000000005e-06, + "loss": 0.1293, + "step": 1278 + }, + { + "epoch": 1.513161786453712, + "grad_norm": 1.5398629903793335, + "learning_rate": 3.4012500000000003e-06, + "loss": 0.1212, + "step": 1279 + }, + { + "epoch": 1.5143448683821354, + "grad_norm": 1.555671215057373, + "learning_rate": 3.4000000000000005e-06, + "loss": 0.1113, + "step": 1280 + }, + { + "epoch": 1.515527950310559, + "grad_norm": 1.738730549812317, + "learning_rate": 3.3987500000000003e-06, + "loss": 0.128, + "step": 1281 + }, + { + "epoch": 1.5167110322389825, + "grad_norm": 1.481830358505249, + "learning_rate": 3.3975e-06, + "loss": 0.1389, + "step": 1282 + }, + { + "epoch": 1.517894114167406, + "grad_norm": 1.7934081554412842, + "learning_rate": 3.3962500000000003e-06, + "loss": 0.1431, + "step": 1283 + }, + { + "epoch": 1.5190771960958296, + "grad_norm": 1.6029057502746582, + "learning_rate": 3.3950000000000005e-06, + "loss": 0.1324, + "step": 1284 + }, + { + "epoch": 1.5202602780242533, + "grad_norm": 1.8634427785873413, + "learning_rate": 3.39375e-06, + "loss": 0.1334, + "step": 1285 + }, + { + "epoch": 1.5214433599526767, + "grad_norm": 1.3210428953170776, + "learning_rate": 3.3925e-06, + "loss": 0.1224, + "step": 1286 + }, + { + "epoch": 1.5226264418811004, + "grad_norm": 1.8886276483535767, + "learning_rate": 3.3912500000000004e-06, + "loss": 0.129, + "step": 1287 + }, + { + "epoch": 1.5238095238095237, + "grad_norm": 1.3924386501312256, + "learning_rate": 3.3900000000000006e-06, + "loss": 0.1277, + "step": 1288 + }, + { + "epoch": 1.5249926057379475, + "grad_norm": 1.4204670190811157, + "learning_rate": 3.38875e-06, + "loss": 0.1232, + "step": 1289 + }, + { + "epoch": 1.5261756876663708, + "grad_norm": 1.5176576375961304, + "learning_rate": 3.3875e-06, + "loss": 0.123, + "step": 1290 + }, + { + "epoch": 1.5273587695947946, + "grad_norm": 1.549462080001831, + "learning_rate": 3.3862500000000004e-06, + "loss": 0.135, + "step": 1291 + }, + { + "epoch": 1.5285418515232179, + "grad_norm": 1.4946926832199097, + "learning_rate": 3.3850000000000006e-06, + "loss": 0.1234, + "step": 1292 + }, + { + "epoch": 1.5297249334516416, + "grad_norm": 1.550441861152649, + "learning_rate": 3.38375e-06, + "loss": 0.1228, + "step": 1293 + }, + { + "epoch": 1.530908015380065, + "grad_norm": 1.6477125883102417, + "learning_rate": 3.3825000000000002e-06, + "loss": 0.1252, + "step": 1294 + }, + { + "epoch": 1.5320910973084887, + "grad_norm": 1.838528037071228, + "learning_rate": 3.3812500000000004e-06, + "loss": 0.1235, + "step": 1295 + }, + { + "epoch": 1.533274179236912, + "grad_norm": 2.020090341567993, + "learning_rate": 3.3800000000000007e-06, + "loss": 0.1397, + "step": 1296 + }, + { + "epoch": 1.5344572611653358, + "grad_norm": 1.5766327381134033, + "learning_rate": 3.37875e-06, + "loss": 0.1295, + "step": 1297 + }, + { + "epoch": 1.5356403430937593, + "grad_norm": 1.4336988925933838, + "learning_rate": 3.3775000000000003e-06, + "loss": 0.1287, + "step": 1298 + }, + { + "epoch": 1.5368234250221828, + "grad_norm": 1.6806628704071045, + "learning_rate": 3.3762500000000005e-06, + "loss": 0.1232, + "step": 1299 + }, + { + "epoch": 1.5380065069506064, + "grad_norm": 1.543700933456421, + "learning_rate": 3.3750000000000003e-06, + "loss": 0.1152, + "step": 1300 + }, + { + "epoch": 1.53918958887903, + "grad_norm": 1.4754140377044678, + "learning_rate": 3.37375e-06, + "loss": 0.127, + "step": 1301 + }, + { + "epoch": 1.5403726708074534, + "grad_norm": 1.4444950819015503, + "learning_rate": 3.3725000000000003e-06, + "loss": 0.1283, + "step": 1302 + }, + { + "epoch": 1.541555752735877, + "grad_norm": 1.8669140338897705, + "learning_rate": 3.3712500000000005e-06, + "loss": 0.1314, + "step": 1303 + }, + { + "epoch": 1.5427388346643005, + "grad_norm": 1.467099905014038, + "learning_rate": 3.3700000000000003e-06, + "loss": 0.1226, + "step": 1304 + }, + { + "epoch": 1.543921916592724, + "grad_norm": 2.098670482635498, + "learning_rate": 3.36875e-06, + "loss": 0.1498, + "step": 1305 + }, + { + "epoch": 1.5451049985211476, + "grad_norm": 1.5185511112213135, + "learning_rate": 3.3675000000000004e-06, + "loss": 0.1208, + "step": 1306 + }, + { + "epoch": 1.5462880804495711, + "grad_norm": 1.4257383346557617, + "learning_rate": 3.36625e-06, + "loss": 0.119, + "step": 1307 + }, + { + "epoch": 1.5474711623779946, + "grad_norm": 1.2572530508041382, + "learning_rate": 3.3650000000000004e-06, + "loss": 0.1126, + "step": 1308 + }, + { + "epoch": 1.5486542443064182, + "grad_norm": 1.5880950689315796, + "learning_rate": 3.36375e-06, + "loss": 0.1339, + "step": 1309 + }, + { + "epoch": 1.5498373262348417, + "grad_norm": 1.7804405689239502, + "learning_rate": 3.3625000000000004e-06, + "loss": 0.1326, + "step": 1310 + }, + { + "epoch": 1.5510204081632653, + "grad_norm": 1.6037368774414062, + "learning_rate": 3.3612500000000002e-06, + "loss": 0.1198, + "step": 1311 + }, + { + "epoch": 1.552203490091689, + "grad_norm": 1.6822205781936646, + "learning_rate": 3.3600000000000004e-06, + "loss": 0.1241, + "step": 1312 + }, + { + "epoch": 1.5533865720201123, + "grad_norm": 1.6317545175552368, + "learning_rate": 3.3587500000000002e-06, + "loss": 0.1257, + "step": 1313 + }, + { + "epoch": 1.554569653948536, + "grad_norm": 1.8221899271011353, + "learning_rate": 3.3575e-06, + "loss": 0.1312, + "step": 1314 + }, + { + "epoch": 1.5557527358769594, + "grad_norm": 1.776476502418518, + "learning_rate": 3.3562500000000003e-06, + "loss": 0.1236, + "step": 1315 + }, + { + "epoch": 1.5569358178053831, + "grad_norm": 1.599238634109497, + "learning_rate": 3.3550000000000005e-06, + "loss": 0.1177, + "step": 1316 + }, + { + "epoch": 1.5581188997338065, + "grad_norm": 2.0098705291748047, + "learning_rate": 3.35375e-06, + "loss": 0.1402, + "step": 1317 + }, + { + "epoch": 1.5593019816622302, + "grad_norm": 1.295106291770935, + "learning_rate": 3.3525e-06, + "loss": 0.1209, + "step": 1318 + }, + { + "epoch": 1.5604850635906535, + "grad_norm": 1.6536020040512085, + "learning_rate": 3.3512500000000003e-06, + "loss": 0.1259, + "step": 1319 + }, + { + "epoch": 1.5616681455190773, + "grad_norm": 1.701365351676941, + "learning_rate": 3.3500000000000005e-06, + "loss": 0.1392, + "step": 1320 + }, + { + "epoch": 1.5628512274475006, + "grad_norm": 1.5777922868728638, + "learning_rate": 3.34875e-06, + "loss": 0.1205, + "step": 1321 + }, + { + "epoch": 1.5640343093759244, + "grad_norm": 1.5953125953674316, + "learning_rate": 3.3475e-06, + "loss": 0.1261, + "step": 1322 + }, + { + "epoch": 1.5652173913043477, + "grad_norm": 1.3697435855865479, + "learning_rate": 3.3462500000000004e-06, + "loss": 0.1234, + "step": 1323 + }, + { + "epoch": 1.5664004732327714, + "grad_norm": 1.5795377492904663, + "learning_rate": 3.3450000000000006e-06, + "loss": 0.1259, + "step": 1324 + }, + { + "epoch": 1.567583555161195, + "grad_norm": 1.4269118309020996, + "learning_rate": 3.34375e-06, + "loss": 0.1293, + "step": 1325 + }, + { + "epoch": 1.5687666370896185, + "grad_norm": 1.4944474697113037, + "learning_rate": 3.3425e-06, + "loss": 0.124, + "step": 1326 + }, + { + "epoch": 1.569949719018042, + "grad_norm": 1.629680871963501, + "learning_rate": 3.3412500000000004e-06, + "loss": 0.1405, + "step": 1327 + }, + { + "epoch": 1.5711328009464656, + "grad_norm": 1.5238053798675537, + "learning_rate": 3.3400000000000006e-06, + "loss": 0.1186, + "step": 1328 + }, + { + "epoch": 1.572315882874889, + "grad_norm": 1.9159058332443237, + "learning_rate": 3.33875e-06, + "loss": 0.1498, + "step": 1329 + }, + { + "epoch": 1.5734989648033126, + "grad_norm": 1.5055044889450073, + "learning_rate": 3.3375000000000002e-06, + "loss": 0.1265, + "step": 1330 + }, + { + "epoch": 1.5746820467317362, + "grad_norm": 1.4665433168411255, + "learning_rate": 3.3362500000000005e-06, + "loss": 0.1249, + "step": 1331 + }, + { + "epoch": 1.5758651286601597, + "grad_norm": 1.486828088760376, + "learning_rate": 3.3350000000000003e-06, + "loss": 0.116, + "step": 1332 + }, + { + "epoch": 1.5770482105885832, + "grad_norm": 1.4532381296157837, + "learning_rate": 3.33375e-06, + "loss": 0.1323, + "step": 1333 + }, + { + "epoch": 1.5782312925170068, + "grad_norm": 1.5230575799942017, + "learning_rate": 3.3325000000000003e-06, + "loss": 0.1303, + "step": 1334 + }, + { + "epoch": 1.5794143744454303, + "grad_norm": 1.5705646276474, + "learning_rate": 3.3312500000000005e-06, + "loss": 0.132, + "step": 1335 + }, + { + "epoch": 1.5805974563738538, + "grad_norm": 1.6271547079086304, + "learning_rate": 3.3300000000000003e-06, + "loss": 0.1273, + "step": 1336 + }, + { + "epoch": 1.5817805383022774, + "grad_norm": 1.684235692024231, + "learning_rate": 3.32875e-06, + "loss": 0.1235, + "step": 1337 + }, + { + "epoch": 1.582963620230701, + "grad_norm": 1.3318500518798828, + "learning_rate": 3.3275000000000003e-06, + "loss": 0.1168, + "step": 1338 + }, + { + "epoch": 1.5841467021591247, + "grad_norm": 1.6092740297317505, + "learning_rate": 3.32625e-06, + "loss": 0.1212, + "step": 1339 + }, + { + "epoch": 1.585329784087548, + "grad_norm": 1.654579758644104, + "learning_rate": 3.3250000000000004e-06, + "loss": 0.1256, + "step": 1340 + }, + { + "epoch": 1.5865128660159717, + "grad_norm": 1.6255793571472168, + "learning_rate": 3.32375e-06, + "loss": 0.1275, + "step": 1341 + }, + { + "epoch": 1.587695947944395, + "grad_norm": 1.6256195306777954, + "learning_rate": 3.3225000000000004e-06, + "loss": 0.129, + "step": 1342 + }, + { + "epoch": 1.5888790298728188, + "grad_norm": 1.561197280883789, + "learning_rate": 3.32125e-06, + "loss": 0.1338, + "step": 1343 + }, + { + "epoch": 1.5900621118012421, + "grad_norm": 1.6948766708374023, + "learning_rate": 3.3200000000000004e-06, + "loss": 0.1284, + "step": 1344 + }, + { + "epoch": 1.5912451937296659, + "grad_norm": 1.6150776147842407, + "learning_rate": 3.31875e-06, + "loss": 0.1233, + "step": 1345 + }, + { + "epoch": 1.5924282756580892, + "grad_norm": 1.5668781995773315, + "learning_rate": 3.3175e-06, + "loss": 0.1206, + "step": 1346 + }, + { + "epoch": 1.593611357586513, + "grad_norm": 1.7716835737228394, + "learning_rate": 3.3162500000000002e-06, + "loss": 0.1449, + "step": 1347 + }, + { + "epoch": 1.5947944395149363, + "grad_norm": 1.8333115577697754, + "learning_rate": 3.3150000000000004e-06, + "loss": 0.1389, + "step": 1348 + }, + { + "epoch": 1.59597752144336, + "grad_norm": 1.5289050340652466, + "learning_rate": 3.31375e-06, + "loss": 0.1255, + "step": 1349 + }, + { + "epoch": 1.5971606033717833, + "grad_norm": 1.4431642293930054, + "learning_rate": 3.3125e-06, + "loss": 0.1243, + "step": 1350 + }, + { + "epoch": 1.598343685300207, + "grad_norm": 1.4966899156570435, + "learning_rate": 3.3112500000000003e-06, + "loss": 0.1341, + "step": 1351 + }, + { + "epoch": 1.5995267672286306, + "grad_norm": 1.6540204286575317, + "learning_rate": 3.3100000000000005e-06, + "loss": 0.1449, + "step": 1352 + }, + { + "epoch": 1.6007098491570542, + "grad_norm": 1.755121111869812, + "learning_rate": 3.30875e-06, + "loss": 0.1336, + "step": 1353 + }, + { + "epoch": 1.6018929310854777, + "grad_norm": 1.6204171180725098, + "learning_rate": 3.3075e-06, + "loss": 0.1343, + "step": 1354 + }, + { + "epoch": 1.6030760130139012, + "grad_norm": 1.4811965227127075, + "learning_rate": 3.3062500000000003e-06, + "loss": 0.1334, + "step": 1355 + }, + { + "epoch": 1.6042590949423248, + "grad_norm": 1.5078394412994385, + "learning_rate": 3.3050000000000005e-06, + "loss": 0.1287, + "step": 1356 + }, + { + "epoch": 1.6054421768707483, + "grad_norm": 1.5438920259475708, + "learning_rate": 3.30375e-06, + "loss": 0.1426, + "step": 1357 + }, + { + "epoch": 1.6066252587991718, + "grad_norm": 1.5858154296875, + "learning_rate": 3.3025e-06, + "loss": 0.1284, + "step": 1358 + }, + { + "epoch": 1.6078083407275954, + "grad_norm": 1.5437064170837402, + "learning_rate": 3.3012500000000004e-06, + "loss": 0.1216, + "step": 1359 + }, + { + "epoch": 1.608991422656019, + "grad_norm": 1.676072120666504, + "learning_rate": 3.3000000000000006e-06, + "loss": 0.1339, + "step": 1360 + }, + { + "epoch": 1.6101745045844424, + "grad_norm": 1.4908864498138428, + "learning_rate": 3.29875e-06, + "loss": 0.1327, + "step": 1361 + }, + { + "epoch": 1.611357586512866, + "grad_norm": 1.3954812288284302, + "learning_rate": 3.2975e-06, + "loss": 0.1208, + "step": 1362 + }, + { + "epoch": 1.6125406684412895, + "grad_norm": 1.5311059951782227, + "learning_rate": 3.2962500000000004e-06, + "loss": 0.1373, + "step": 1363 + }, + { + "epoch": 1.613723750369713, + "grad_norm": 1.729414463043213, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.1332, + "step": 1364 + }, + { + "epoch": 1.6149068322981366, + "grad_norm": 1.4659161567687988, + "learning_rate": 3.29375e-06, + "loss": 0.1255, + "step": 1365 + }, + { + "epoch": 1.6160899142265603, + "grad_norm": 1.4245327711105347, + "learning_rate": 3.2925000000000002e-06, + "loss": 0.1197, + "step": 1366 + }, + { + "epoch": 1.6172729961549837, + "grad_norm": 1.7252216339111328, + "learning_rate": 3.2912500000000005e-06, + "loss": 0.1338, + "step": 1367 + }, + { + "epoch": 1.6184560780834074, + "grad_norm": 1.9947254657745361, + "learning_rate": 3.2900000000000003e-06, + "loss": 0.1351, + "step": 1368 + }, + { + "epoch": 1.6196391600118307, + "grad_norm": 1.4900641441345215, + "learning_rate": 3.28875e-06, + "loss": 0.1224, + "step": 1369 + }, + { + "epoch": 1.6208222419402545, + "grad_norm": 1.8801724910736084, + "learning_rate": 3.2875000000000003e-06, + "loss": 0.1403, + "step": 1370 + }, + { + "epoch": 1.6220053238686778, + "grad_norm": 1.6036713123321533, + "learning_rate": 3.28625e-06, + "loss": 0.1386, + "step": 1371 + }, + { + "epoch": 1.6231884057971016, + "grad_norm": 1.87173330783844, + "learning_rate": 3.2850000000000003e-06, + "loss": 0.1221, + "step": 1372 + }, + { + "epoch": 1.6243714877255249, + "grad_norm": 1.4788272380828857, + "learning_rate": 3.28375e-06, + "loss": 0.1209, + "step": 1373 + }, + { + "epoch": 1.6255545696539486, + "grad_norm": 1.476340651512146, + "learning_rate": 3.2825000000000003e-06, + "loss": 0.1269, + "step": 1374 + }, + { + "epoch": 1.626737651582372, + "grad_norm": 1.5400975942611694, + "learning_rate": 3.28125e-06, + "loss": 0.1285, + "step": 1375 + }, + { + "epoch": 1.6279207335107957, + "grad_norm": 1.6609598398208618, + "learning_rate": 3.2800000000000004e-06, + "loss": 0.1258, + "step": 1376 + }, + { + "epoch": 1.629103815439219, + "grad_norm": 1.3722655773162842, + "learning_rate": 3.2787500000000006e-06, + "loss": 0.1302, + "step": 1377 + }, + { + "epoch": 1.6302868973676428, + "grad_norm": 1.3829176425933838, + "learning_rate": 3.2775e-06, + "loss": 0.1304, + "step": 1378 + }, + { + "epoch": 1.6314699792960663, + "grad_norm": 1.8069695234298706, + "learning_rate": 3.27625e-06, + "loss": 0.1163, + "step": 1379 + }, + { + "epoch": 1.6326530612244898, + "grad_norm": 2.075058937072754, + "learning_rate": 3.2750000000000004e-06, + "loss": 0.1313, + "step": 1380 + }, + { + "epoch": 1.6338361431529134, + "grad_norm": 1.5353800058364868, + "learning_rate": 3.2737500000000006e-06, + "loss": 0.1321, + "step": 1381 + }, + { + "epoch": 1.635019225081337, + "grad_norm": 1.525162935256958, + "learning_rate": 3.2725e-06, + "loss": 0.1219, + "step": 1382 + }, + { + "epoch": 1.6362023070097604, + "grad_norm": 1.74387526512146, + "learning_rate": 3.2712500000000002e-06, + "loss": 0.1298, + "step": 1383 + }, + { + "epoch": 1.637385388938184, + "grad_norm": 1.5597354173660278, + "learning_rate": 3.2700000000000005e-06, + "loss": 0.1272, + "step": 1384 + }, + { + "epoch": 1.6385684708666075, + "grad_norm": 1.5195435285568237, + "learning_rate": 3.2687500000000007e-06, + "loss": 0.1232, + "step": 1385 + }, + { + "epoch": 1.639751552795031, + "grad_norm": 1.5236088037490845, + "learning_rate": 3.2675e-06, + "loss": 0.1322, + "step": 1386 + }, + { + "epoch": 1.6409346347234546, + "grad_norm": 1.3826849460601807, + "learning_rate": 3.2662500000000003e-06, + "loss": 0.1233, + "step": 1387 + }, + { + "epoch": 1.6421177166518781, + "grad_norm": 1.6898058652877808, + "learning_rate": 3.2650000000000005e-06, + "loss": 0.1351, + "step": 1388 + }, + { + "epoch": 1.6433007985803016, + "grad_norm": 1.4246114492416382, + "learning_rate": 3.2637500000000007e-06, + "loss": 0.1291, + "step": 1389 + }, + { + "epoch": 1.6444838805087252, + "grad_norm": 1.6492928266525269, + "learning_rate": 3.2625e-06, + "loss": 0.1311, + "step": 1390 + }, + { + "epoch": 1.6456669624371487, + "grad_norm": 1.9392977952957153, + "learning_rate": 3.2612500000000003e-06, + "loss": 0.1355, + "step": 1391 + }, + { + "epoch": 1.6468500443655723, + "grad_norm": 1.518025517463684, + "learning_rate": 3.2600000000000006e-06, + "loss": 0.1333, + "step": 1392 + }, + { + "epoch": 1.648033126293996, + "grad_norm": 1.3588926792144775, + "learning_rate": 3.2587500000000004e-06, + "loss": 0.1254, + "step": 1393 + }, + { + "epoch": 1.6492162082224193, + "grad_norm": 1.3850055932998657, + "learning_rate": 3.2575e-06, + "loss": 0.1242, + "step": 1394 + }, + { + "epoch": 1.650399290150843, + "grad_norm": 1.6214162111282349, + "learning_rate": 3.2562500000000004e-06, + "loss": 0.1329, + "step": 1395 + }, + { + "epoch": 1.6515823720792664, + "grad_norm": 1.5566637516021729, + "learning_rate": 3.255e-06, + "loss": 0.126, + "step": 1396 + }, + { + "epoch": 1.6527654540076901, + "grad_norm": 1.455531358718872, + "learning_rate": 3.2537500000000004e-06, + "loss": 0.123, + "step": 1397 + }, + { + "epoch": 1.6539485359361135, + "grad_norm": 1.6385713815689087, + "learning_rate": 3.2525e-06, + "loss": 0.1243, + "step": 1398 + }, + { + "epoch": 1.6551316178645372, + "grad_norm": 1.5812796354293823, + "learning_rate": 3.2512500000000004e-06, + "loss": 0.1304, + "step": 1399 + }, + { + "epoch": 1.6563146997929605, + "grad_norm": 1.4917150735855103, + "learning_rate": 3.2500000000000002e-06, + "loss": 0.114, + "step": 1400 + }, + { + "epoch": 1.6574977817213843, + "grad_norm": 1.6036486625671387, + "learning_rate": 3.2487500000000004e-06, + "loss": 0.1296, + "step": 1401 + }, + { + "epoch": 1.6586808636498076, + "grad_norm": 1.8148932456970215, + "learning_rate": 3.2475000000000002e-06, + "loss": 0.1407, + "step": 1402 + }, + { + "epoch": 1.6598639455782314, + "grad_norm": 1.3486355543136597, + "learning_rate": 3.24625e-06, + "loss": 0.1262, + "step": 1403 + }, + { + "epoch": 1.6610470275066547, + "grad_norm": 1.47530996799469, + "learning_rate": 3.2450000000000003e-06, + "loss": 0.1233, + "step": 1404 + }, + { + "epoch": 1.6622301094350784, + "grad_norm": 1.6789849996566772, + "learning_rate": 3.2437500000000005e-06, + "loss": 0.1326, + "step": 1405 + }, + { + "epoch": 1.663413191363502, + "grad_norm": 1.4145750999450684, + "learning_rate": 3.2425000000000003e-06, + "loss": 0.13, + "step": 1406 + }, + { + "epoch": 1.6645962732919255, + "grad_norm": 1.5003387928009033, + "learning_rate": 3.24125e-06, + "loss": 0.1239, + "step": 1407 + }, + { + "epoch": 1.665779355220349, + "grad_norm": 1.4523297548294067, + "learning_rate": 3.2400000000000003e-06, + "loss": 0.1223, + "step": 1408 + }, + { + "epoch": 1.6669624371487726, + "grad_norm": 1.7971453666687012, + "learning_rate": 3.2387500000000005e-06, + "loss": 0.1248, + "step": 1409 + }, + { + "epoch": 1.668145519077196, + "grad_norm": 1.575683355331421, + "learning_rate": 3.2375e-06, + "loss": 0.1354, + "step": 1410 + }, + { + "epoch": 1.6693286010056196, + "grad_norm": 1.594149112701416, + "learning_rate": 3.23625e-06, + "loss": 0.131, + "step": 1411 + }, + { + "epoch": 1.6705116829340432, + "grad_norm": 1.573073148727417, + "learning_rate": 3.2350000000000004e-06, + "loss": 0.1261, + "step": 1412 + }, + { + "epoch": 1.6716947648624667, + "grad_norm": 1.6821691989898682, + "learning_rate": 3.2337500000000006e-06, + "loss": 0.1219, + "step": 1413 + }, + { + "epoch": 1.6728778467908902, + "grad_norm": 1.6000694036483765, + "learning_rate": 3.2325e-06, + "loss": 0.1252, + "step": 1414 + }, + { + "epoch": 1.6740609287193138, + "grad_norm": 1.667115330696106, + "learning_rate": 3.23125e-06, + "loss": 0.1371, + "step": 1415 + }, + { + "epoch": 1.6752440106477373, + "grad_norm": 1.7219778299331665, + "learning_rate": 3.2300000000000004e-06, + "loss": 0.1265, + "step": 1416 + }, + { + "epoch": 1.6764270925761608, + "grad_norm": 1.7029204368591309, + "learning_rate": 3.2287500000000006e-06, + "loss": 0.1322, + "step": 1417 + }, + { + "epoch": 1.6776101745045846, + "grad_norm": 1.562144160270691, + "learning_rate": 3.2275e-06, + "loss": 0.1388, + "step": 1418 + }, + { + "epoch": 1.678793256433008, + "grad_norm": 1.6732985973358154, + "learning_rate": 3.2262500000000002e-06, + "loss": 0.1282, + "step": 1419 + }, + { + "epoch": 1.6799763383614317, + "grad_norm": 1.3361759185791016, + "learning_rate": 3.2250000000000005e-06, + "loss": 0.1165, + "step": 1420 + }, + { + "epoch": 1.681159420289855, + "grad_norm": 1.439334750175476, + "learning_rate": 3.2237500000000007e-06, + "loss": 0.1364, + "step": 1421 + }, + { + "epoch": 1.6823425022182787, + "grad_norm": 1.414253830909729, + "learning_rate": 3.2225e-06, + "loss": 0.1294, + "step": 1422 + }, + { + "epoch": 1.683525584146702, + "grad_norm": 1.32809317111969, + "learning_rate": 3.2212500000000003e-06, + "loss": 0.125, + "step": 1423 + }, + { + "epoch": 1.6847086660751258, + "grad_norm": 1.639321208000183, + "learning_rate": 3.2200000000000005e-06, + "loss": 0.1341, + "step": 1424 + }, + { + "epoch": 1.6858917480035491, + "grad_norm": 1.4606908559799194, + "learning_rate": 3.2187500000000003e-06, + "loss": 0.1301, + "step": 1425 + }, + { + "epoch": 1.6870748299319729, + "grad_norm": 1.5274593830108643, + "learning_rate": 3.2175e-06, + "loss": 0.1179, + "step": 1426 + }, + { + "epoch": 1.6882579118603962, + "grad_norm": 1.3704708814620972, + "learning_rate": 3.2162500000000003e-06, + "loss": 0.1326, + "step": 1427 + }, + { + "epoch": 1.68944099378882, + "grad_norm": 1.5484576225280762, + "learning_rate": 3.215e-06, + "loss": 0.1221, + "step": 1428 + }, + { + "epoch": 1.6906240757172433, + "grad_norm": 1.291685938835144, + "learning_rate": 3.2137500000000004e-06, + "loss": 0.1206, + "step": 1429 + }, + { + "epoch": 1.691807157645667, + "grad_norm": 1.5275603532791138, + "learning_rate": 3.2125e-06, + "loss": 0.1246, + "step": 1430 + }, + { + "epoch": 1.6929902395740906, + "grad_norm": 1.6574949026107788, + "learning_rate": 3.2112500000000004e-06, + "loss": 0.1332, + "step": 1431 + }, + { + "epoch": 1.694173321502514, + "grad_norm": 1.4093583822250366, + "learning_rate": 3.21e-06, + "loss": 0.1313, + "step": 1432 + }, + { + "epoch": 1.6953564034309376, + "grad_norm": 1.8137526512145996, + "learning_rate": 3.2087500000000004e-06, + "loss": 0.1346, + "step": 1433 + }, + { + "epoch": 1.6965394853593612, + "grad_norm": 1.5104016065597534, + "learning_rate": 3.2075e-06, + "loss": 0.1287, + "step": 1434 + }, + { + "epoch": 1.6977225672877847, + "grad_norm": 1.700255036354065, + "learning_rate": 3.20625e-06, + "loss": 0.1207, + "step": 1435 + }, + { + "epoch": 1.6989056492162082, + "grad_norm": 1.3068333864212036, + "learning_rate": 3.2050000000000002e-06, + "loss": 0.1264, + "step": 1436 + }, + { + "epoch": 1.7000887311446318, + "grad_norm": 1.6368564367294312, + "learning_rate": 3.2037500000000005e-06, + "loss": 0.1218, + "step": 1437 + }, + { + "epoch": 1.7012718130730553, + "grad_norm": 1.9458857774734497, + "learning_rate": 3.2025000000000003e-06, + "loss": 0.1374, + "step": 1438 + }, + { + "epoch": 1.7024548950014788, + "grad_norm": 1.493390679359436, + "learning_rate": 3.20125e-06, + "loss": 0.119, + "step": 1439 + }, + { + "epoch": 1.7036379769299024, + "grad_norm": 1.732709527015686, + "learning_rate": 3.2000000000000003e-06, + "loss": 0.1332, + "step": 1440 + }, + { + "epoch": 1.704821058858326, + "grad_norm": 1.5783692598342896, + "learning_rate": 3.1987500000000005e-06, + "loss": 0.1312, + "step": 1441 + }, + { + "epoch": 1.7060041407867494, + "grad_norm": 1.627793788909912, + "learning_rate": 3.1975e-06, + "loss": 0.1322, + "step": 1442 + }, + { + "epoch": 1.707187222715173, + "grad_norm": 1.5680112838745117, + "learning_rate": 3.19625e-06, + "loss": 0.1307, + "step": 1443 + }, + { + "epoch": 1.7083703046435965, + "grad_norm": 1.964853286743164, + "learning_rate": 3.1950000000000003e-06, + "loss": 0.1322, + "step": 1444 + }, + { + "epoch": 1.7095533865720203, + "grad_norm": 1.7595957517623901, + "learning_rate": 3.1937500000000006e-06, + "loss": 0.1285, + "step": 1445 + }, + { + "epoch": 1.7107364685004436, + "grad_norm": 1.409660816192627, + "learning_rate": 3.1925e-06, + "loss": 0.1317, + "step": 1446 + }, + { + "epoch": 1.7119195504288673, + "grad_norm": 1.42058265209198, + "learning_rate": 3.19125e-06, + "loss": 0.1311, + "step": 1447 + }, + { + "epoch": 1.7131026323572907, + "grad_norm": 1.427865982055664, + "learning_rate": 3.1900000000000004e-06, + "loss": 0.1363, + "step": 1448 + }, + { + "epoch": 1.7142857142857144, + "grad_norm": 1.5116851329803467, + "learning_rate": 3.1887500000000006e-06, + "loss": 0.1302, + "step": 1449 + }, + { + "epoch": 1.7154687962141377, + "grad_norm": 1.638986349105835, + "learning_rate": 3.1875e-06, + "loss": 0.1406, + "step": 1450 + }, + { + "epoch": 1.7166518781425615, + "grad_norm": 1.340059757232666, + "learning_rate": 3.18625e-06, + "loss": 0.1211, + "step": 1451 + }, + { + "epoch": 1.7178349600709848, + "grad_norm": 1.349758267402649, + "learning_rate": 3.1850000000000004e-06, + "loss": 0.1291, + "step": 1452 + }, + { + "epoch": 1.7190180419994086, + "grad_norm": 1.7397477626800537, + "learning_rate": 3.1837500000000006e-06, + "loss": 0.1315, + "step": 1453 + }, + { + "epoch": 1.7202011239278319, + "grad_norm": 1.460517168045044, + "learning_rate": 3.1825e-06, + "loss": 0.131, + "step": 1454 + }, + { + "epoch": 1.7213842058562556, + "grad_norm": 1.4067931175231934, + "learning_rate": 3.1812500000000002e-06, + "loss": 0.118, + "step": 1455 + }, + { + "epoch": 1.722567287784679, + "grad_norm": 1.739151120185852, + "learning_rate": 3.1800000000000005e-06, + "loss": 0.1299, + "step": 1456 + }, + { + "epoch": 1.7237503697131027, + "grad_norm": 1.7023221254348755, + "learning_rate": 3.1787500000000003e-06, + "loss": 0.135, + "step": 1457 + }, + { + "epoch": 1.7249334516415262, + "grad_norm": 1.8165796995162964, + "learning_rate": 3.1775e-06, + "loss": 0.1308, + "step": 1458 + }, + { + "epoch": 1.7261165335699498, + "grad_norm": 1.8294546604156494, + "learning_rate": 3.1762500000000003e-06, + "loss": 0.1264, + "step": 1459 + }, + { + "epoch": 1.7272996154983733, + "grad_norm": 1.7271509170532227, + "learning_rate": 3.175e-06, + "loss": 0.1412, + "step": 1460 + }, + { + "epoch": 1.7284826974267968, + "grad_norm": 2.020429849624634, + "learning_rate": 3.1737500000000003e-06, + "loss": 0.1223, + "step": 1461 + }, + { + "epoch": 1.7296657793552204, + "grad_norm": 1.5288043022155762, + "learning_rate": 3.1725e-06, + "loss": 0.1323, + "step": 1462 + }, + { + "epoch": 1.730848861283644, + "grad_norm": 1.562862753868103, + "learning_rate": 3.1712500000000003e-06, + "loss": 0.1273, + "step": 1463 + }, + { + "epoch": 1.7320319432120674, + "grad_norm": 1.463289737701416, + "learning_rate": 3.17e-06, + "loss": 0.1165, + "step": 1464 + }, + { + "epoch": 1.733215025140491, + "grad_norm": 1.3559701442718506, + "learning_rate": 3.1687500000000004e-06, + "loss": 0.1259, + "step": 1465 + }, + { + "epoch": 1.7343981070689145, + "grad_norm": 1.5908234119415283, + "learning_rate": 3.1675e-06, + "loss": 0.1319, + "step": 1466 + }, + { + "epoch": 1.735581188997338, + "grad_norm": 1.513815999031067, + "learning_rate": 3.16625e-06, + "loss": 0.1294, + "step": 1467 + }, + { + "epoch": 1.7367642709257616, + "grad_norm": 1.5134055614471436, + "learning_rate": 3.165e-06, + "loss": 0.1325, + "step": 1468 + }, + { + "epoch": 1.7379473528541851, + "grad_norm": 1.6640589237213135, + "learning_rate": 3.1637500000000004e-06, + "loss": 0.1312, + "step": 1469 + }, + { + "epoch": 1.7391304347826086, + "grad_norm": 1.4721251726150513, + "learning_rate": 3.1625000000000002e-06, + "loss": 0.1306, + "step": 1470 + }, + { + "epoch": 1.7403135167110322, + "grad_norm": 1.541951060295105, + "learning_rate": 3.16125e-06, + "loss": 0.1264, + "step": 1471 + }, + { + "epoch": 1.741496598639456, + "grad_norm": 1.5553547143936157, + "learning_rate": 3.1600000000000002e-06, + "loss": 0.1317, + "step": 1472 + }, + { + "epoch": 1.7426796805678793, + "grad_norm": 1.707440733909607, + "learning_rate": 3.1587500000000005e-06, + "loss": 0.125, + "step": 1473 + }, + { + "epoch": 1.743862762496303, + "grad_norm": 1.8376456499099731, + "learning_rate": 3.1575e-06, + "loss": 0.13, + "step": 1474 + }, + { + "epoch": 1.7450458444247263, + "grad_norm": 1.4374293088912964, + "learning_rate": 3.15625e-06, + "loss": 0.1235, + "step": 1475 + }, + { + "epoch": 1.74622892635315, + "grad_norm": 1.4540596008300781, + "learning_rate": 3.1550000000000003e-06, + "loss": 0.1239, + "step": 1476 + }, + { + "epoch": 1.7474120082815734, + "grad_norm": 1.6718724966049194, + "learning_rate": 3.1537500000000005e-06, + "loss": 0.1355, + "step": 1477 + }, + { + "epoch": 1.7485950902099971, + "grad_norm": 1.4354175329208374, + "learning_rate": 3.1525e-06, + "loss": 0.1329, + "step": 1478 + }, + { + "epoch": 1.7497781721384205, + "grad_norm": 1.7448992729187012, + "learning_rate": 3.15125e-06, + "loss": 0.1252, + "step": 1479 + }, + { + "epoch": 1.7509612540668442, + "grad_norm": 1.4661272764205933, + "learning_rate": 3.1500000000000003e-06, + "loss": 0.1232, + "step": 1480 + }, + { + "epoch": 1.7521443359952675, + "grad_norm": 1.3992351293563843, + "learning_rate": 3.1487500000000006e-06, + "loss": 0.1185, + "step": 1481 + }, + { + "epoch": 1.7533274179236913, + "grad_norm": 1.6304161548614502, + "learning_rate": 3.1475e-06, + "loss": 0.1219, + "step": 1482 + }, + { + "epoch": 1.7545104998521146, + "grad_norm": 1.5025243759155273, + "learning_rate": 3.14625e-06, + "loss": 0.1282, + "step": 1483 + }, + { + "epoch": 1.7556935817805384, + "grad_norm": 2.1731865406036377, + "learning_rate": 3.1450000000000004e-06, + "loss": 0.1186, + "step": 1484 + }, + { + "epoch": 1.756876663708962, + "grad_norm": 1.5010864734649658, + "learning_rate": 3.1437500000000006e-06, + "loss": 0.1138, + "step": 1485 + }, + { + "epoch": 1.7580597456373854, + "grad_norm": 1.4763137102127075, + "learning_rate": 3.1425e-06, + "loss": 0.1125, + "step": 1486 + }, + { + "epoch": 1.759242827565809, + "grad_norm": 1.489912509918213, + "learning_rate": 3.14125e-06, + "loss": 0.1148, + "step": 1487 + }, + { + "epoch": 1.7604259094942325, + "grad_norm": 1.49180269241333, + "learning_rate": 3.1400000000000004e-06, + "loss": 0.1352, + "step": 1488 + }, + { + "epoch": 1.761608991422656, + "grad_norm": 1.4568650722503662, + "learning_rate": 3.1387500000000002e-06, + "loss": 0.1217, + "step": 1489 + }, + { + "epoch": 1.7627920733510796, + "grad_norm": 1.2612231969833374, + "learning_rate": 3.1375e-06, + "loss": 0.1116, + "step": 1490 + }, + { + "epoch": 1.763975155279503, + "grad_norm": 1.4808948040008545, + "learning_rate": 3.1362500000000003e-06, + "loss": 0.1291, + "step": 1491 + }, + { + "epoch": 1.7651582372079266, + "grad_norm": 1.6505444049835205, + "learning_rate": 3.135e-06, + "loss": 0.1359, + "step": 1492 + }, + { + "epoch": 1.7663413191363502, + "grad_norm": 1.4132620096206665, + "learning_rate": 3.1337500000000003e-06, + "loss": 0.1328, + "step": 1493 + }, + { + "epoch": 1.7675244010647737, + "grad_norm": 1.3438763618469238, + "learning_rate": 3.1325e-06, + "loss": 0.1195, + "step": 1494 + }, + { + "epoch": 1.7687074829931972, + "grad_norm": 1.788949728012085, + "learning_rate": 3.1312500000000003e-06, + "loss": 0.1281, + "step": 1495 + }, + { + "epoch": 1.7698905649216208, + "grad_norm": 1.563748836517334, + "learning_rate": 3.13e-06, + "loss": 0.1224, + "step": 1496 + }, + { + "epoch": 1.7710736468500443, + "grad_norm": 1.5647658109664917, + "learning_rate": 3.1287500000000003e-06, + "loss": 0.1292, + "step": 1497 + }, + { + "epoch": 1.7722567287784678, + "grad_norm": 1.4464058876037598, + "learning_rate": 3.1275e-06, + "loss": 0.1245, + "step": 1498 + }, + { + "epoch": 1.7734398107068916, + "grad_norm": 1.7593613862991333, + "learning_rate": 3.12625e-06, + "loss": 0.142, + "step": 1499 + }, + { + "epoch": 1.774622892635315, + "grad_norm": 1.5450142621994019, + "learning_rate": 3.125e-06, + "loss": 0.1318, + "step": 1500 + }, + { + "epoch": 1.7758059745637387, + "grad_norm": 1.4954546689987183, + "learning_rate": 3.1237500000000004e-06, + "loss": 0.1274, + "step": 1501 + }, + { + "epoch": 1.776989056492162, + "grad_norm": 1.7929115295410156, + "learning_rate": 3.1225000000000006e-06, + "loss": 0.1393, + "step": 1502 + }, + { + "epoch": 1.7781721384205857, + "grad_norm": 1.659194827079773, + "learning_rate": 3.12125e-06, + "loss": 0.1297, + "step": 1503 + }, + { + "epoch": 1.779355220349009, + "grad_norm": 1.3256151676177979, + "learning_rate": 3.12e-06, + "loss": 0.124, + "step": 1504 + }, + { + "epoch": 1.7805383022774328, + "grad_norm": 1.7122669219970703, + "learning_rate": 3.1187500000000004e-06, + "loss": 0.1224, + "step": 1505 + }, + { + "epoch": 1.7817213842058561, + "grad_norm": 1.6573967933654785, + "learning_rate": 3.1175000000000006e-06, + "loss": 0.1244, + "step": 1506 + }, + { + "epoch": 1.7829044661342799, + "grad_norm": 1.4729957580566406, + "learning_rate": 3.11625e-06, + "loss": 0.1231, + "step": 1507 + }, + { + "epoch": 1.7840875480627032, + "grad_norm": 1.5873345136642456, + "learning_rate": 3.1150000000000002e-06, + "loss": 0.1284, + "step": 1508 + }, + { + "epoch": 1.785270629991127, + "grad_norm": 1.7143833637237549, + "learning_rate": 3.1137500000000005e-06, + "loss": 0.1239, + "step": 1509 + }, + { + "epoch": 1.7864537119195503, + "grad_norm": 1.4365202188491821, + "learning_rate": 3.1125000000000007e-06, + "loss": 0.1286, + "step": 1510 + }, + { + "epoch": 1.787636793847974, + "grad_norm": 1.3060005903244019, + "learning_rate": 3.11125e-06, + "loss": 0.1307, + "step": 1511 + }, + { + "epoch": 1.7888198757763976, + "grad_norm": 1.9179496765136719, + "learning_rate": 3.1100000000000003e-06, + "loss": 0.1321, + "step": 1512 + }, + { + "epoch": 1.790002957704821, + "grad_norm": 1.4626022577285767, + "learning_rate": 3.1087500000000005e-06, + "loss": 0.1314, + "step": 1513 + }, + { + "epoch": 1.7911860396332446, + "grad_norm": 1.3617677688598633, + "learning_rate": 3.1075000000000003e-06, + "loss": 0.1338, + "step": 1514 + }, + { + "epoch": 1.7923691215616682, + "grad_norm": 1.4807071685791016, + "learning_rate": 3.10625e-06, + "loss": 0.1299, + "step": 1515 + }, + { + "epoch": 1.7935522034900917, + "grad_norm": 1.5578906536102295, + "learning_rate": 3.1050000000000003e-06, + "loss": 0.1306, + "step": 1516 + }, + { + "epoch": 1.7947352854185152, + "grad_norm": 1.7090046405792236, + "learning_rate": 3.1037500000000006e-06, + "loss": 0.1172, + "step": 1517 + }, + { + "epoch": 1.7959183673469388, + "grad_norm": 1.5331132411956787, + "learning_rate": 3.1025000000000004e-06, + "loss": 0.1242, + "step": 1518 + }, + { + "epoch": 1.7971014492753623, + "grad_norm": 1.6458816528320312, + "learning_rate": 3.10125e-06, + "loss": 0.1418, + "step": 1519 + }, + { + "epoch": 1.7982845312037858, + "grad_norm": 1.5989110469818115, + "learning_rate": 3.1000000000000004e-06, + "loss": 0.1398, + "step": 1520 + }, + { + "epoch": 1.7994676131322094, + "grad_norm": 1.3849536180496216, + "learning_rate": 3.09875e-06, + "loss": 0.1062, + "step": 1521 + }, + { + "epoch": 1.800650695060633, + "grad_norm": 1.4989910125732422, + "learning_rate": 3.0975000000000004e-06, + "loss": 0.1186, + "step": 1522 + }, + { + "epoch": 1.8018337769890564, + "grad_norm": 1.7686622142791748, + "learning_rate": 3.0962500000000002e-06, + "loss": 0.1308, + "step": 1523 + }, + { + "epoch": 1.80301685891748, + "grad_norm": 1.5845211744308472, + "learning_rate": 3.0950000000000004e-06, + "loss": 0.1405, + "step": 1524 + }, + { + "epoch": 1.8041999408459035, + "grad_norm": 1.4296889305114746, + "learning_rate": 3.0937500000000002e-06, + "loss": 0.1233, + "step": 1525 + }, + { + "epoch": 1.8053830227743273, + "grad_norm": 1.473272442817688, + "learning_rate": 3.0925000000000005e-06, + "loss": 0.1193, + "step": 1526 + }, + { + "epoch": 1.8065661047027506, + "grad_norm": 1.6567400693893433, + "learning_rate": 3.0912500000000003e-06, + "loss": 0.1297, + "step": 1527 + }, + { + "epoch": 1.8077491866311743, + "grad_norm": 1.6100224256515503, + "learning_rate": 3.09e-06, + "loss": 0.1212, + "step": 1528 + }, + { + "epoch": 1.8089322685595977, + "grad_norm": 1.918100357055664, + "learning_rate": 3.0887500000000003e-06, + "loss": 0.1236, + "step": 1529 + }, + { + "epoch": 1.8101153504880214, + "grad_norm": 1.9279192686080933, + "learning_rate": 3.0875000000000005e-06, + "loss": 0.1248, + "step": 1530 + }, + { + "epoch": 1.8112984324164447, + "grad_norm": 1.5977054834365845, + "learning_rate": 3.08625e-06, + "loss": 0.1253, + "step": 1531 + }, + { + "epoch": 1.8124815143448685, + "grad_norm": 1.5097906589508057, + "learning_rate": 3.085e-06, + "loss": 0.1329, + "step": 1532 + }, + { + "epoch": 1.8136645962732918, + "grad_norm": 1.5370922088623047, + "learning_rate": 3.0837500000000003e-06, + "loss": 0.1309, + "step": 1533 + }, + { + "epoch": 1.8148476782017156, + "grad_norm": 1.6556071043014526, + "learning_rate": 3.0825000000000006e-06, + "loss": 0.1431, + "step": 1534 + }, + { + "epoch": 1.8160307601301389, + "grad_norm": 1.910179853439331, + "learning_rate": 3.08125e-06, + "loss": 0.1275, + "step": 1535 + }, + { + "epoch": 1.8172138420585626, + "grad_norm": 2.0411887168884277, + "learning_rate": 3.08e-06, + "loss": 0.1381, + "step": 1536 + }, + { + "epoch": 1.818396923986986, + "grad_norm": 1.3288357257843018, + "learning_rate": 3.0787500000000004e-06, + "loss": 0.1288, + "step": 1537 + }, + { + "epoch": 1.8195800059154097, + "grad_norm": 1.36674165725708, + "learning_rate": 3.0775000000000006e-06, + "loss": 0.1233, + "step": 1538 + }, + { + "epoch": 1.8207630878438332, + "grad_norm": 1.6524626016616821, + "learning_rate": 3.07625e-06, + "loss": 0.1364, + "step": 1539 + }, + { + "epoch": 1.8219461697722568, + "grad_norm": 1.3931044340133667, + "learning_rate": 3.075e-06, + "loss": 0.1228, + "step": 1540 + }, + { + "epoch": 1.8231292517006803, + "grad_norm": 1.5236079692840576, + "learning_rate": 3.0737500000000004e-06, + "loss": 0.1264, + "step": 1541 + }, + { + "epoch": 1.8243123336291038, + "grad_norm": 1.4967103004455566, + "learning_rate": 3.0725000000000007e-06, + "loss": 0.1253, + "step": 1542 + }, + { + "epoch": 1.8254954155575274, + "grad_norm": 1.7943062782287598, + "learning_rate": 3.07125e-06, + "loss": 0.1454, + "step": 1543 + }, + { + "epoch": 1.826678497485951, + "grad_norm": 1.9367706775665283, + "learning_rate": 3.0700000000000003e-06, + "loss": 0.1295, + "step": 1544 + }, + { + "epoch": 1.8278615794143744, + "grad_norm": 1.6375234127044678, + "learning_rate": 3.0687500000000005e-06, + "loss": 0.1338, + "step": 1545 + }, + { + "epoch": 1.829044661342798, + "grad_norm": 1.2888541221618652, + "learning_rate": 3.0675000000000003e-06, + "loss": 0.1208, + "step": 1546 + }, + { + "epoch": 1.8302277432712215, + "grad_norm": 1.4628701210021973, + "learning_rate": 3.06625e-06, + "loss": 0.1283, + "step": 1547 + }, + { + "epoch": 1.831410825199645, + "grad_norm": 1.3949475288391113, + "learning_rate": 3.0650000000000003e-06, + "loss": 0.1136, + "step": 1548 + }, + { + "epoch": 1.8325939071280686, + "grad_norm": 1.352097511291504, + "learning_rate": 3.0637500000000005e-06, + "loss": 0.1135, + "step": 1549 + }, + { + "epoch": 1.8337769890564921, + "grad_norm": 1.6653172969818115, + "learning_rate": 3.0625000000000003e-06, + "loss": 0.1231, + "step": 1550 + }, + { + "epoch": 1.8349600709849159, + "grad_norm": 1.6152822971343994, + "learning_rate": 3.06125e-06, + "loss": 0.1381, + "step": 1551 + }, + { + "epoch": 1.8361431529133392, + "grad_norm": 1.4173654317855835, + "learning_rate": 3.0600000000000003e-06, + "loss": 0.1238, + "step": 1552 + }, + { + "epoch": 1.837326234841763, + "grad_norm": 1.737138032913208, + "learning_rate": 3.05875e-06, + "loss": 0.1263, + "step": 1553 + }, + { + "epoch": 1.8385093167701863, + "grad_norm": 1.7955886125564575, + "learning_rate": 3.0575000000000004e-06, + "loss": 0.1295, + "step": 1554 + }, + { + "epoch": 1.83969239869861, + "grad_norm": 1.739182949066162, + "learning_rate": 3.05625e-06, + "loss": 0.1405, + "step": 1555 + }, + { + "epoch": 1.8408754806270333, + "grad_norm": 1.3528324365615845, + "learning_rate": 3.0550000000000004e-06, + "loss": 0.1237, + "step": 1556 + }, + { + "epoch": 1.842058562555457, + "grad_norm": 1.5265575647354126, + "learning_rate": 3.05375e-06, + "loss": 0.1367, + "step": 1557 + }, + { + "epoch": 1.8432416444838804, + "grad_norm": 1.7132514715194702, + "learning_rate": 3.0525000000000004e-06, + "loss": 0.1482, + "step": 1558 + }, + { + "epoch": 1.8444247264123041, + "grad_norm": 1.4762369394302368, + "learning_rate": 3.0512500000000002e-06, + "loss": 0.1365, + "step": 1559 + }, + { + "epoch": 1.8456078083407275, + "grad_norm": 1.5722016096115112, + "learning_rate": 3.05e-06, + "loss": 0.1422, + "step": 1560 + }, + { + "epoch": 1.8467908902691512, + "grad_norm": 1.3664965629577637, + "learning_rate": 3.0487500000000002e-06, + "loss": 0.1271, + "step": 1561 + }, + { + "epoch": 1.8479739721975745, + "grad_norm": 1.524817705154419, + "learning_rate": 3.0475000000000005e-06, + "loss": 0.1165, + "step": 1562 + }, + { + "epoch": 1.8491570541259983, + "grad_norm": 1.6021283864974976, + "learning_rate": 3.04625e-06, + "loss": 0.1284, + "step": 1563 + }, + { + "epoch": 1.8503401360544216, + "grad_norm": 1.6855648756027222, + "learning_rate": 3.045e-06, + "loss": 0.1355, + "step": 1564 + }, + { + "epoch": 1.8515232179828454, + "grad_norm": 1.5778870582580566, + "learning_rate": 3.0437500000000003e-06, + "loss": 0.1318, + "step": 1565 + }, + { + "epoch": 1.852706299911269, + "grad_norm": 1.748910665512085, + "learning_rate": 3.0425000000000005e-06, + "loss": 0.1324, + "step": 1566 + }, + { + "epoch": 1.8538893818396924, + "grad_norm": 1.7581230401992798, + "learning_rate": 3.04125e-06, + "loss": 0.1261, + "step": 1567 + }, + { + "epoch": 1.855072463768116, + "grad_norm": 1.5424503087997437, + "learning_rate": 3.04e-06, + "loss": 0.1185, + "step": 1568 + }, + { + "epoch": 1.8562555456965395, + "grad_norm": 1.7005378007888794, + "learning_rate": 3.0387500000000003e-06, + "loss": 0.1173, + "step": 1569 + }, + { + "epoch": 1.857438627624963, + "grad_norm": 1.4160120487213135, + "learning_rate": 3.0375000000000006e-06, + "loss": 0.1267, + "step": 1570 + }, + { + "epoch": 1.8586217095533866, + "grad_norm": 1.4962800741195679, + "learning_rate": 3.03625e-06, + "loss": 0.1261, + "step": 1571 + }, + { + "epoch": 1.85980479148181, + "grad_norm": 1.833624005317688, + "learning_rate": 3.035e-06, + "loss": 0.1365, + "step": 1572 + }, + { + "epoch": 1.8609878734102336, + "grad_norm": 1.7055097818374634, + "learning_rate": 3.0337500000000004e-06, + "loss": 0.1255, + "step": 1573 + }, + { + "epoch": 1.8621709553386572, + "grad_norm": 1.634921669960022, + "learning_rate": 3.0325000000000006e-06, + "loss": 0.1279, + "step": 1574 + }, + { + "epoch": 1.8633540372670807, + "grad_norm": 1.5820231437683105, + "learning_rate": 3.03125e-06, + "loss": 0.1249, + "step": 1575 + }, + { + "epoch": 1.8645371191955042, + "grad_norm": 2.064409017562866, + "learning_rate": 3.0300000000000002e-06, + "loss": 0.1284, + "step": 1576 + }, + { + "epoch": 1.8657202011239278, + "grad_norm": 1.557623267173767, + "learning_rate": 3.0287500000000004e-06, + "loss": 0.1255, + "step": 1577 + }, + { + "epoch": 1.8669032830523515, + "grad_norm": 1.5268746614456177, + "learning_rate": 3.0275000000000002e-06, + "loss": 0.1335, + "step": 1578 + }, + { + "epoch": 1.8680863649807748, + "grad_norm": 1.7602530717849731, + "learning_rate": 3.02625e-06, + "loss": 0.1381, + "step": 1579 + }, + { + "epoch": 1.8692694469091986, + "grad_norm": 1.6634483337402344, + "learning_rate": 3.0250000000000003e-06, + "loss": 0.1266, + "step": 1580 + }, + { + "epoch": 1.870452528837622, + "grad_norm": 1.617326259613037, + "learning_rate": 3.0237500000000005e-06, + "loss": 0.1174, + "step": 1581 + }, + { + "epoch": 1.8716356107660457, + "grad_norm": 1.6061689853668213, + "learning_rate": 3.0225000000000003e-06, + "loss": 0.1222, + "step": 1582 + }, + { + "epoch": 1.872818692694469, + "grad_norm": 1.5044060945510864, + "learning_rate": 3.02125e-06, + "loss": 0.1188, + "step": 1583 + }, + { + "epoch": 1.8740017746228927, + "grad_norm": 1.6560814380645752, + "learning_rate": 3.0200000000000003e-06, + "loss": 0.1254, + "step": 1584 + }, + { + "epoch": 1.875184856551316, + "grad_norm": 1.8804951906204224, + "learning_rate": 3.01875e-06, + "loss": 0.1406, + "step": 1585 + }, + { + "epoch": 1.8763679384797398, + "grad_norm": 1.6935606002807617, + "learning_rate": 3.0175000000000003e-06, + "loss": 0.1206, + "step": 1586 + }, + { + "epoch": 1.8775510204081631, + "grad_norm": 1.4293185472488403, + "learning_rate": 3.01625e-06, + "loss": 0.1207, + "step": 1587 + }, + { + "epoch": 1.8787341023365869, + "grad_norm": 1.6188381910324097, + "learning_rate": 3.0150000000000004e-06, + "loss": 0.1244, + "step": 1588 + }, + { + "epoch": 1.8799171842650102, + "grad_norm": 1.5349304676055908, + "learning_rate": 3.01375e-06, + "loss": 0.1357, + "step": 1589 + }, + { + "epoch": 1.881100266193434, + "grad_norm": 1.493005633354187, + "learning_rate": 3.0125000000000004e-06, + "loss": 0.1224, + "step": 1590 + }, + { + "epoch": 1.8822833481218575, + "grad_norm": 1.378043532371521, + "learning_rate": 3.01125e-06, + "loss": 0.119, + "step": 1591 + }, + { + "epoch": 1.883466430050281, + "grad_norm": 1.6582642793655396, + "learning_rate": 3.01e-06, + "loss": 0.1409, + "step": 1592 + }, + { + "epoch": 1.8846495119787046, + "grad_norm": 1.4895769357681274, + "learning_rate": 3.00875e-06, + "loss": 0.1348, + "step": 1593 + }, + { + "epoch": 1.885832593907128, + "grad_norm": 1.497365117073059, + "learning_rate": 3.0075000000000004e-06, + "loss": 0.1278, + "step": 1594 + }, + { + "epoch": 1.8870156758355516, + "grad_norm": 1.6291561126708984, + "learning_rate": 3.00625e-06, + "loss": 0.1302, + "step": 1595 + }, + { + "epoch": 1.8881987577639752, + "grad_norm": 1.6288950443267822, + "learning_rate": 3.005e-06, + "loss": 0.126, + "step": 1596 + }, + { + "epoch": 1.8893818396923987, + "grad_norm": 1.7120941877365112, + "learning_rate": 3.0037500000000003e-06, + "loss": 0.1401, + "step": 1597 + }, + { + "epoch": 1.8905649216208222, + "grad_norm": 1.374051809310913, + "learning_rate": 3.0025000000000005e-06, + "loss": 0.1224, + "step": 1598 + }, + { + "epoch": 1.8917480035492458, + "grad_norm": 1.7910175323486328, + "learning_rate": 3.00125e-06, + "loss": 0.1245, + "step": 1599 + }, + { + "epoch": 1.8929310854776693, + "grad_norm": 1.703108310699463, + "learning_rate": 3e-06, + "loss": 0.127, + "step": 1600 + }, + { + "epoch": 1.8941141674060928, + "grad_norm": 1.713757872581482, + "learning_rate": 2.9987500000000003e-06, + "loss": 0.1252, + "step": 1601 + }, + { + "epoch": 1.8952972493345164, + "grad_norm": 1.3957245349884033, + "learning_rate": 2.9975000000000005e-06, + "loss": 0.1177, + "step": 1602 + }, + { + "epoch": 1.89648033126294, + "grad_norm": 1.5342751741409302, + "learning_rate": 2.99625e-06, + "loss": 0.1289, + "step": 1603 + }, + { + "epoch": 1.8976634131913634, + "grad_norm": 1.918247103691101, + "learning_rate": 2.995e-06, + "loss": 0.1311, + "step": 1604 + }, + { + "epoch": 1.8988464951197872, + "grad_norm": 1.6796822547912598, + "learning_rate": 2.9937500000000003e-06, + "loss": 0.1379, + "step": 1605 + }, + { + "epoch": 1.9000295770482105, + "grad_norm": 1.6882449388504028, + "learning_rate": 2.9925000000000006e-06, + "loss": 0.1239, + "step": 1606 + }, + { + "epoch": 1.9012126589766343, + "grad_norm": 1.5194429159164429, + "learning_rate": 2.99125e-06, + "loss": 0.1268, + "step": 1607 + }, + { + "epoch": 1.9023957409050576, + "grad_norm": 1.9256019592285156, + "learning_rate": 2.99e-06, + "loss": 0.131, + "step": 1608 + }, + { + "epoch": 1.9035788228334813, + "grad_norm": 1.3945752382278442, + "learning_rate": 2.9887500000000004e-06, + "loss": 0.1308, + "step": 1609 + }, + { + "epoch": 1.9047619047619047, + "grad_norm": 1.3421820402145386, + "learning_rate": 2.9875e-06, + "loss": 0.1265, + "step": 1610 + }, + { + "epoch": 1.9059449866903284, + "grad_norm": 1.6099538803100586, + "learning_rate": 2.98625e-06, + "loss": 0.1303, + "step": 1611 + }, + { + "epoch": 1.9071280686187517, + "grad_norm": 1.584062933921814, + "learning_rate": 2.9850000000000002e-06, + "loss": 0.1275, + "step": 1612 + }, + { + "epoch": 1.9083111505471755, + "grad_norm": 1.5516650676727295, + "learning_rate": 2.9837500000000004e-06, + "loss": 0.115, + "step": 1613 + }, + { + "epoch": 1.9094942324755988, + "grad_norm": 1.5857435464859009, + "learning_rate": 2.9825000000000002e-06, + "loss": 0.1206, + "step": 1614 + }, + { + "epoch": 1.9106773144040226, + "grad_norm": 1.7921782732009888, + "learning_rate": 2.98125e-06, + "loss": 0.1418, + "step": 1615 + }, + { + "epoch": 1.9118603963324459, + "grad_norm": 1.6154537200927734, + "learning_rate": 2.9800000000000003e-06, + "loss": 0.1414, + "step": 1616 + }, + { + "epoch": 1.9130434782608696, + "grad_norm": 1.69894540309906, + "learning_rate": 2.97875e-06, + "loss": 0.1311, + "step": 1617 + }, + { + "epoch": 1.9142265601892932, + "grad_norm": 1.6527644395828247, + "learning_rate": 2.9775000000000003e-06, + "loss": 0.1342, + "step": 1618 + }, + { + "epoch": 1.9154096421177167, + "grad_norm": 1.2943997383117676, + "learning_rate": 2.97625e-06, + "loss": 0.1315, + "step": 1619 + }, + { + "epoch": 1.9165927240461402, + "grad_norm": 1.4852967262268066, + "learning_rate": 2.9750000000000003e-06, + "loss": 0.1168, + "step": 1620 + }, + { + "epoch": 1.9177758059745638, + "grad_norm": 1.597418189048767, + "learning_rate": 2.97375e-06, + "loss": 0.1256, + "step": 1621 + }, + { + "epoch": 1.9189588879029873, + "grad_norm": 1.4352481365203857, + "learning_rate": 2.9725000000000003e-06, + "loss": 0.1345, + "step": 1622 + }, + { + "epoch": 1.9201419698314108, + "grad_norm": 2.0390994548797607, + "learning_rate": 2.97125e-06, + "loss": 0.1445, + "step": 1623 + }, + { + "epoch": 1.9213250517598344, + "grad_norm": 1.4389827251434326, + "learning_rate": 2.97e-06, + "loss": 0.1188, + "step": 1624 + }, + { + "epoch": 1.922508133688258, + "grad_norm": 2.0218381881713867, + "learning_rate": 2.96875e-06, + "loss": 0.1331, + "step": 1625 + }, + { + "epoch": 1.9236912156166814, + "grad_norm": 1.4945030212402344, + "learning_rate": 2.9675000000000004e-06, + "loss": 0.1317, + "step": 1626 + }, + { + "epoch": 1.924874297545105, + "grad_norm": 1.5888441801071167, + "learning_rate": 2.9662500000000006e-06, + "loss": 0.1225, + "step": 1627 + }, + { + "epoch": 1.9260573794735285, + "grad_norm": 1.4204803705215454, + "learning_rate": 2.965e-06, + "loss": 0.1323, + "step": 1628 + }, + { + "epoch": 1.927240461401952, + "grad_norm": 1.4633713960647583, + "learning_rate": 2.96375e-06, + "loss": 0.1176, + "step": 1629 + }, + { + "epoch": 1.9284235433303756, + "grad_norm": 1.8409489393234253, + "learning_rate": 2.9625000000000004e-06, + "loss": 0.1176, + "step": 1630 + }, + { + "epoch": 1.9296066252587991, + "grad_norm": 1.4938435554504395, + "learning_rate": 2.9612500000000007e-06, + "loss": 0.1255, + "step": 1631 + }, + { + "epoch": 1.9307897071872229, + "grad_norm": 1.4316428899765015, + "learning_rate": 2.96e-06, + "loss": 0.122, + "step": 1632 + }, + { + "epoch": 1.9319727891156462, + "grad_norm": 1.8549553155899048, + "learning_rate": 2.9587500000000003e-06, + "loss": 0.1216, + "step": 1633 + }, + { + "epoch": 1.93315587104407, + "grad_norm": 1.577225923538208, + "learning_rate": 2.9575000000000005e-06, + "loss": 0.1227, + "step": 1634 + }, + { + "epoch": 1.9343389529724933, + "grad_norm": 1.6939531564712524, + "learning_rate": 2.9562500000000007e-06, + "loss": 0.1407, + "step": 1635 + }, + { + "epoch": 1.935522034900917, + "grad_norm": 1.6354894638061523, + "learning_rate": 2.955e-06, + "loss": 0.1485, + "step": 1636 + }, + { + "epoch": 1.9367051168293403, + "grad_norm": 1.400924563407898, + "learning_rate": 2.9537500000000003e-06, + "loss": 0.1262, + "step": 1637 + }, + { + "epoch": 1.937888198757764, + "grad_norm": 1.8341224193572998, + "learning_rate": 2.9525000000000005e-06, + "loss": 0.133, + "step": 1638 + }, + { + "epoch": 1.9390712806861874, + "grad_norm": 1.6949177980422974, + "learning_rate": 2.9512500000000003e-06, + "loss": 0.1333, + "step": 1639 + }, + { + "epoch": 1.9402543626146111, + "grad_norm": 1.516874074935913, + "learning_rate": 2.95e-06, + "loss": 0.1276, + "step": 1640 + }, + { + "epoch": 1.9414374445430345, + "grad_norm": 1.2863750457763672, + "learning_rate": 2.9487500000000004e-06, + "loss": 0.1193, + "step": 1641 + }, + { + "epoch": 1.9426205264714582, + "grad_norm": 1.6023945808410645, + "learning_rate": 2.9475e-06, + "loss": 0.1287, + "step": 1642 + }, + { + "epoch": 1.9438036083998815, + "grad_norm": 1.514906406402588, + "learning_rate": 2.9462500000000004e-06, + "loss": 0.1223, + "step": 1643 + }, + { + "epoch": 1.9449866903283053, + "grad_norm": 1.7632625102996826, + "learning_rate": 2.945e-06, + "loss": 0.1351, + "step": 1644 + }, + { + "epoch": 1.9461697722567288, + "grad_norm": 2.0445871353149414, + "learning_rate": 2.9437500000000004e-06, + "loss": 0.124, + "step": 1645 + }, + { + "epoch": 1.9473528541851524, + "grad_norm": 1.615757942199707, + "learning_rate": 2.9425e-06, + "loss": 0.1312, + "step": 1646 + }, + { + "epoch": 1.948535936113576, + "grad_norm": 1.5237065553665161, + "learning_rate": 2.9412500000000004e-06, + "loss": 0.1319, + "step": 1647 + }, + { + "epoch": 1.9497190180419994, + "grad_norm": 1.4474513530731201, + "learning_rate": 2.9400000000000002e-06, + "loss": 0.1213, + "step": 1648 + }, + { + "epoch": 1.950902099970423, + "grad_norm": 1.7680572271347046, + "learning_rate": 2.93875e-06, + "loss": 0.1257, + "step": 1649 + }, + { + "epoch": 1.9520851818988465, + "grad_norm": 1.5007939338684082, + "learning_rate": 2.9375000000000003e-06, + "loss": 0.1239, + "step": 1650 + }, + { + "epoch": 1.95326826382727, + "grad_norm": 1.370653510093689, + "learning_rate": 2.9362500000000005e-06, + "loss": 0.1272, + "step": 1651 + }, + { + "epoch": 1.9544513457556936, + "grad_norm": 1.825180172920227, + "learning_rate": 2.9350000000000003e-06, + "loss": 0.1304, + "step": 1652 + }, + { + "epoch": 1.955634427684117, + "grad_norm": 1.506386399269104, + "learning_rate": 2.93375e-06, + "loss": 0.1381, + "step": 1653 + }, + { + "epoch": 1.9568175096125406, + "grad_norm": 1.3242523670196533, + "learning_rate": 2.9325000000000003e-06, + "loss": 0.132, + "step": 1654 + }, + { + "epoch": 1.9580005915409642, + "grad_norm": 1.323806881904602, + "learning_rate": 2.9312500000000005e-06, + "loss": 0.1266, + "step": 1655 + }, + { + "epoch": 1.9591836734693877, + "grad_norm": 1.7909647226333618, + "learning_rate": 2.93e-06, + "loss": 0.1379, + "step": 1656 + }, + { + "epoch": 1.9603667553978112, + "grad_norm": 1.9128189086914062, + "learning_rate": 2.92875e-06, + "loss": 0.1446, + "step": 1657 + }, + { + "epoch": 1.9615498373262348, + "grad_norm": 1.6031441688537598, + "learning_rate": 2.9275000000000003e-06, + "loss": 0.1358, + "step": 1658 + }, + { + "epoch": 1.9627329192546585, + "grad_norm": 1.6824264526367188, + "learning_rate": 2.9262500000000006e-06, + "loss": 0.1299, + "step": 1659 + }, + { + "epoch": 1.9639160011830818, + "grad_norm": 1.903107762336731, + "learning_rate": 2.925e-06, + "loss": 0.1273, + "step": 1660 + }, + { + "epoch": 1.9650990831115056, + "grad_norm": 1.3546808958053589, + "learning_rate": 2.92375e-06, + "loss": 0.126, + "step": 1661 + }, + { + "epoch": 1.966282165039929, + "grad_norm": 1.785469651222229, + "learning_rate": 2.9225000000000004e-06, + "loss": 0.1343, + "step": 1662 + }, + { + "epoch": 1.9674652469683527, + "grad_norm": 1.4613838195800781, + "learning_rate": 2.9212500000000006e-06, + "loss": 0.1274, + "step": 1663 + }, + { + "epoch": 1.968648328896776, + "grad_norm": 1.971160650253296, + "learning_rate": 2.92e-06, + "loss": 0.1294, + "step": 1664 + }, + { + "epoch": 1.9698314108251997, + "grad_norm": 1.5510122776031494, + "learning_rate": 2.9187500000000002e-06, + "loss": 0.1172, + "step": 1665 + }, + { + "epoch": 1.971014492753623, + "grad_norm": 1.426884651184082, + "learning_rate": 2.9175000000000004e-06, + "loss": 0.127, + "step": 1666 + }, + { + "epoch": 1.9721975746820468, + "grad_norm": 2.032762289047241, + "learning_rate": 2.9162500000000007e-06, + "loss": 0.1373, + "step": 1667 + }, + { + "epoch": 1.9733806566104701, + "grad_norm": 1.5049248933792114, + "learning_rate": 2.915e-06, + "loss": 0.1248, + "step": 1668 + }, + { + "epoch": 1.9745637385388939, + "grad_norm": 1.4161971807479858, + "learning_rate": 2.9137500000000003e-06, + "loss": 0.1307, + "step": 1669 + }, + { + "epoch": 1.9757468204673172, + "grad_norm": 1.657818078994751, + "learning_rate": 2.9125000000000005e-06, + "loss": 0.1293, + "step": 1670 + }, + { + "epoch": 1.976929902395741, + "grad_norm": 1.5728939771652222, + "learning_rate": 2.9112500000000003e-06, + "loss": 0.1272, + "step": 1671 + }, + { + "epoch": 1.9781129843241645, + "grad_norm": 1.4363834857940674, + "learning_rate": 2.91e-06, + "loss": 0.1259, + "step": 1672 + }, + { + "epoch": 1.979296066252588, + "grad_norm": 1.968100666999817, + "learning_rate": 2.9087500000000003e-06, + "loss": 0.1286, + "step": 1673 + }, + { + "epoch": 1.9804791481810116, + "grad_norm": 1.4426119327545166, + "learning_rate": 2.9075e-06, + "loss": 0.1134, + "step": 1674 + }, + { + "epoch": 1.981662230109435, + "grad_norm": 1.4456043243408203, + "learning_rate": 2.9062500000000003e-06, + "loss": 0.1222, + "step": 1675 + }, + { + "epoch": 1.9828453120378586, + "grad_norm": 1.7659021615982056, + "learning_rate": 2.905e-06, + "loss": 0.1387, + "step": 1676 + }, + { + "epoch": 1.9840283939662822, + "grad_norm": 1.4951870441436768, + "learning_rate": 2.9037500000000004e-06, + "loss": 0.1255, + "step": 1677 + }, + { + "epoch": 1.9852114758947057, + "grad_norm": 1.3990627527236938, + "learning_rate": 2.9025e-06, + "loss": 0.1392, + "step": 1678 + }, + { + "epoch": 1.9863945578231292, + "grad_norm": 1.4861100912094116, + "learning_rate": 2.9012500000000004e-06, + "loss": 0.125, + "step": 1679 + }, + { + "epoch": 1.9875776397515528, + "grad_norm": 1.5938737392425537, + "learning_rate": 2.9e-06, + "loss": 0.1309, + "step": 1680 + }, + { + "epoch": 1.9887607216799763, + "grad_norm": 1.7330044507980347, + "learning_rate": 2.89875e-06, + "loss": 0.1379, + "step": 1681 + }, + { + "epoch": 1.9899438036083998, + "grad_norm": 1.6108825206756592, + "learning_rate": 2.8975e-06, + "loss": 0.1406, + "step": 1682 + }, + { + "epoch": 1.9911268855368234, + "grad_norm": 1.7151660919189453, + "learning_rate": 2.8962500000000004e-06, + "loss": 0.1271, + "step": 1683 + }, + { + "epoch": 1.9923099674652471, + "grad_norm": 1.5955464839935303, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.117, + "step": 1684 + }, + { + "epoch": 1.9934930493936704, + "grad_norm": 1.5395134687423706, + "learning_rate": 2.89375e-06, + "loss": 0.1229, + "step": 1685 + }, + { + "epoch": 1.9946761313220942, + "grad_norm": 1.6514579057693481, + "learning_rate": 2.8925000000000003e-06, + "loss": 0.1193, + "step": 1686 + }, + { + "epoch": 1.9958592132505175, + "grad_norm": 1.590994954109192, + "learning_rate": 2.8912500000000005e-06, + "loss": 0.1452, + "step": 1687 + }, + { + "epoch": 1.9970422951789413, + "grad_norm": 1.7981655597686768, + "learning_rate": 2.89e-06, + "loss": 0.1357, + "step": 1688 + }, + { + "epoch": 1.9982253771073646, + "grad_norm": 1.3664666414260864, + "learning_rate": 2.88875e-06, + "loss": 0.1257, + "step": 1689 + }, + { + "epoch": 1.9994084590357883, + "grad_norm": 1.444676160812378, + "learning_rate": 2.8875000000000003e-06, + "loss": 0.1145, + "step": 1690 + }, + { + "epoch": 2.0005915409642117, + "grad_norm": 1.228034257888794, + "learning_rate": 2.8862500000000005e-06, + "loss": 0.0938, + "step": 1691 + }, + { + "epoch": 2.0017746228926354, + "grad_norm": 1.3616929054260254, + "learning_rate": 2.885e-06, + "loss": 0.0805, + "step": 1692 + }, + { + "epoch": 2.0029577048210587, + "grad_norm": 1.2058523893356323, + "learning_rate": 2.88375e-06, + "loss": 0.0636, + "step": 1693 + }, + { + "epoch": 2.0041407867494825, + "grad_norm": 1.6216148138046265, + "learning_rate": 2.8825000000000004e-06, + "loss": 0.0716, + "step": 1694 + }, + { + "epoch": 2.005323868677906, + "grad_norm": 1.7139812707901, + "learning_rate": 2.8812500000000006e-06, + "loss": 0.0699, + "step": 1695 + }, + { + "epoch": 2.0065069506063296, + "grad_norm": 1.7922900915145874, + "learning_rate": 2.88e-06, + "loss": 0.0743, + "step": 1696 + }, + { + "epoch": 2.007690032534753, + "grad_norm": 1.3808839321136475, + "learning_rate": 2.87875e-06, + "loss": 0.0721, + "step": 1697 + }, + { + "epoch": 2.0088731144631766, + "grad_norm": 1.5058773756027222, + "learning_rate": 2.8775000000000004e-06, + "loss": 0.0653, + "step": 1698 + }, + { + "epoch": 2.0100561963916, + "grad_norm": 1.3363828659057617, + "learning_rate": 2.8762500000000006e-06, + "loss": 0.0726, + "step": 1699 + }, + { + "epoch": 2.0112392783200237, + "grad_norm": 1.1657459735870361, + "learning_rate": 2.875e-06, + "loss": 0.0678, + "step": 1700 + }, + { + "epoch": 2.012422360248447, + "grad_norm": 1.5029836893081665, + "learning_rate": 2.8737500000000002e-06, + "loss": 0.0694, + "step": 1701 + }, + { + "epoch": 2.0136054421768708, + "grad_norm": 1.3520056009292603, + "learning_rate": 2.8725000000000004e-06, + "loss": 0.073, + "step": 1702 + }, + { + "epoch": 2.014788524105294, + "grad_norm": 1.492955207824707, + "learning_rate": 2.8712500000000002e-06, + "loss": 0.0725, + "step": 1703 + }, + { + "epoch": 2.015971606033718, + "grad_norm": 1.5257261991500854, + "learning_rate": 2.87e-06, + "loss": 0.0659, + "step": 1704 + }, + { + "epoch": 2.0171546879621416, + "grad_norm": 1.6145250797271729, + "learning_rate": 2.8687500000000003e-06, + "loss": 0.0692, + "step": 1705 + }, + { + "epoch": 2.018337769890565, + "grad_norm": 1.6256119012832642, + "learning_rate": 2.8675e-06, + "loss": 0.072, + "step": 1706 + }, + { + "epoch": 2.0195208518189887, + "grad_norm": 1.9999773502349854, + "learning_rate": 2.8662500000000003e-06, + "loss": 0.0766, + "step": 1707 + }, + { + "epoch": 2.020703933747412, + "grad_norm": 1.503339409828186, + "learning_rate": 2.865e-06, + "loss": 0.0666, + "step": 1708 + }, + { + "epoch": 2.0218870156758357, + "grad_norm": 1.56365966796875, + "learning_rate": 2.8637500000000003e-06, + "loss": 0.0734, + "step": 1709 + }, + { + "epoch": 2.023070097604259, + "grad_norm": 1.4735122919082642, + "learning_rate": 2.8625e-06, + "loss": 0.0646, + "step": 1710 + }, + { + "epoch": 2.024253179532683, + "grad_norm": 1.7530722618103027, + "learning_rate": 2.8612500000000003e-06, + "loss": 0.0668, + "step": 1711 + }, + { + "epoch": 2.025436261461106, + "grad_norm": 1.7714377641677856, + "learning_rate": 2.86e-06, + "loss": 0.0724, + "step": 1712 + }, + { + "epoch": 2.02661934338953, + "grad_norm": 1.8207966089248657, + "learning_rate": 2.85875e-06, + "loss": 0.0682, + "step": 1713 + }, + { + "epoch": 2.027802425317953, + "grad_norm": 1.5810556411743164, + "learning_rate": 2.8575e-06, + "loss": 0.067, + "step": 1714 + }, + { + "epoch": 2.028985507246377, + "grad_norm": 1.3022476434707642, + "learning_rate": 2.8562500000000004e-06, + "loss": 0.0644, + "step": 1715 + }, + { + "epoch": 2.0301685891748003, + "grad_norm": 1.6061739921569824, + "learning_rate": 2.855e-06, + "loss": 0.079, + "step": 1716 + }, + { + "epoch": 2.031351671103224, + "grad_norm": 1.5881919860839844, + "learning_rate": 2.85375e-06, + "loss": 0.0697, + "step": 1717 + }, + { + "epoch": 2.0325347530316473, + "grad_norm": 1.5229166746139526, + "learning_rate": 2.8525000000000002e-06, + "loss": 0.0684, + "step": 1718 + }, + { + "epoch": 2.033717834960071, + "grad_norm": 1.5570378303527832, + "learning_rate": 2.8512500000000004e-06, + "loss": 0.0737, + "step": 1719 + }, + { + "epoch": 2.0349009168884944, + "grad_norm": 1.605955719947815, + "learning_rate": 2.85e-06, + "loss": 0.0667, + "step": 1720 + }, + { + "epoch": 2.036083998816918, + "grad_norm": 1.4247697591781616, + "learning_rate": 2.84875e-06, + "loss": 0.0673, + "step": 1721 + }, + { + "epoch": 2.0372670807453415, + "grad_norm": 1.4566044807434082, + "learning_rate": 2.8475000000000003e-06, + "loss": 0.0668, + "step": 1722 + }, + { + "epoch": 2.038450162673765, + "grad_norm": 1.6094003915786743, + "learning_rate": 2.8462500000000005e-06, + "loss": 0.0701, + "step": 1723 + }, + { + "epoch": 2.0396332446021885, + "grad_norm": 1.3823623657226562, + "learning_rate": 2.845e-06, + "loss": 0.0654, + "step": 1724 + }, + { + "epoch": 2.0408163265306123, + "grad_norm": 1.5007805824279785, + "learning_rate": 2.84375e-06, + "loss": 0.0658, + "step": 1725 + }, + { + "epoch": 2.0419994084590356, + "grad_norm": 2.096440315246582, + "learning_rate": 2.8425000000000003e-06, + "loss": 0.0598, + "step": 1726 + }, + { + "epoch": 2.0431824903874594, + "grad_norm": 1.4194096326828003, + "learning_rate": 2.8412500000000005e-06, + "loss": 0.0667, + "step": 1727 + }, + { + "epoch": 2.0443655723158827, + "grad_norm": 1.5854614973068237, + "learning_rate": 2.84e-06, + "loss": 0.0691, + "step": 1728 + }, + { + "epoch": 2.0455486542443064, + "grad_norm": 1.6034616231918335, + "learning_rate": 2.83875e-06, + "loss": 0.0719, + "step": 1729 + }, + { + "epoch": 2.0467317361727297, + "grad_norm": 1.7773518562316895, + "learning_rate": 2.8375000000000004e-06, + "loss": 0.0688, + "step": 1730 + }, + { + "epoch": 2.0479148181011535, + "grad_norm": 1.5614265203475952, + "learning_rate": 2.8362500000000006e-06, + "loss": 0.0629, + "step": 1731 + }, + { + "epoch": 2.0490979000295773, + "grad_norm": 1.6351709365844727, + "learning_rate": 2.835e-06, + "loss": 0.075, + "step": 1732 + }, + { + "epoch": 2.0502809819580006, + "grad_norm": 1.4305167198181152, + "learning_rate": 2.83375e-06, + "loss": 0.0661, + "step": 1733 + }, + { + "epoch": 2.0514640638864243, + "grad_norm": 1.65324068069458, + "learning_rate": 2.8325000000000004e-06, + "loss": 0.082, + "step": 1734 + }, + { + "epoch": 2.0526471458148476, + "grad_norm": 1.520620346069336, + "learning_rate": 2.83125e-06, + "loss": 0.0671, + "step": 1735 + }, + { + "epoch": 2.0538302277432714, + "grad_norm": 1.4426178932189941, + "learning_rate": 2.83e-06, + "loss": 0.0591, + "step": 1736 + }, + { + "epoch": 2.0550133096716947, + "grad_norm": 1.5228745937347412, + "learning_rate": 2.8287500000000002e-06, + "loss": 0.0689, + "step": 1737 + }, + { + "epoch": 2.0561963916001185, + "grad_norm": 1.460740327835083, + "learning_rate": 2.8275e-06, + "loss": 0.0634, + "step": 1738 + }, + { + "epoch": 2.057379473528542, + "grad_norm": 1.4124741554260254, + "learning_rate": 2.8262500000000003e-06, + "loss": 0.0693, + "step": 1739 + }, + { + "epoch": 2.0585625554569655, + "grad_norm": 1.3417432308197021, + "learning_rate": 2.825e-06, + "loss": 0.0647, + "step": 1740 + }, + { + "epoch": 2.059745637385389, + "grad_norm": 1.3747636079788208, + "learning_rate": 2.8237500000000003e-06, + "loss": 0.066, + "step": 1741 + }, + { + "epoch": 2.0609287193138126, + "grad_norm": 1.3846276998519897, + "learning_rate": 2.8225e-06, + "loss": 0.0679, + "step": 1742 + }, + { + "epoch": 2.062111801242236, + "grad_norm": 1.6705420017242432, + "learning_rate": 2.8212500000000003e-06, + "loss": 0.0717, + "step": 1743 + }, + { + "epoch": 2.0632948831706597, + "grad_norm": 1.6954606771469116, + "learning_rate": 2.82e-06, + "loss": 0.0712, + "step": 1744 + }, + { + "epoch": 2.064477965099083, + "grad_norm": 1.5552653074264526, + "learning_rate": 2.81875e-06, + "loss": 0.0658, + "step": 1745 + }, + { + "epoch": 2.0656610470275067, + "grad_norm": 1.3747080564498901, + "learning_rate": 2.8175e-06, + "loss": 0.0611, + "step": 1746 + }, + { + "epoch": 2.06684412895593, + "grad_norm": 1.7922300100326538, + "learning_rate": 2.8162500000000004e-06, + "loss": 0.072, + "step": 1747 + }, + { + "epoch": 2.068027210884354, + "grad_norm": 1.8153066635131836, + "learning_rate": 2.815e-06, + "loss": 0.0766, + "step": 1748 + }, + { + "epoch": 2.069210292812777, + "grad_norm": 1.5740113258361816, + "learning_rate": 2.81375e-06, + "loss": 0.0714, + "step": 1749 + }, + { + "epoch": 2.070393374741201, + "grad_norm": 1.5816446542739868, + "learning_rate": 2.8125e-06, + "loss": 0.0711, + "step": 1750 + }, + { + "epoch": 2.071576456669624, + "grad_norm": 1.7444502115249634, + "learning_rate": 2.8112500000000004e-06, + "loss": 0.0674, + "step": 1751 + }, + { + "epoch": 2.072759538598048, + "grad_norm": 1.7014094591140747, + "learning_rate": 2.8100000000000006e-06, + "loss": 0.0716, + "step": 1752 + }, + { + "epoch": 2.0739426205264713, + "grad_norm": 1.6440811157226562, + "learning_rate": 2.80875e-06, + "loss": 0.0669, + "step": 1753 + }, + { + "epoch": 2.075125702454895, + "grad_norm": 1.3924832344055176, + "learning_rate": 2.8075000000000002e-06, + "loss": 0.0717, + "step": 1754 + }, + { + "epoch": 2.0763087843833183, + "grad_norm": 1.7758090496063232, + "learning_rate": 2.8062500000000004e-06, + "loss": 0.0758, + "step": 1755 + }, + { + "epoch": 2.077491866311742, + "grad_norm": 1.444021224975586, + "learning_rate": 2.8050000000000007e-06, + "loss": 0.0642, + "step": 1756 + }, + { + "epoch": 2.078674948240166, + "grad_norm": 1.6287686824798584, + "learning_rate": 2.80375e-06, + "loss": 0.0693, + "step": 1757 + }, + { + "epoch": 2.079858030168589, + "grad_norm": 1.5643075704574585, + "learning_rate": 2.8025000000000003e-06, + "loss": 0.0637, + "step": 1758 + }, + { + "epoch": 2.081041112097013, + "grad_norm": 1.4985655546188354, + "learning_rate": 2.8012500000000005e-06, + "loss": 0.0665, + "step": 1759 + }, + { + "epoch": 2.0822241940254362, + "grad_norm": 1.6045409440994263, + "learning_rate": 2.8000000000000003e-06, + "loss": 0.0782, + "step": 1760 + }, + { + "epoch": 2.08340727595386, + "grad_norm": 1.7768481969833374, + "learning_rate": 2.79875e-06, + "loss": 0.0806, + "step": 1761 + }, + { + "epoch": 2.0845903578822833, + "grad_norm": 1.667496681213379, + "learning_rate": 2.7975000000000003e-06, + "loss": 0.0747, + "step": 1762 + }, + { + "epoch": 2.085773439810707, + "grad_norm": 1.6415412425994873, + "learning_rate": 2.7962500000000005e-06, + "loss": 0.0695, + "step": 1763 + }, + { + "epoch": 2.0869565217391304, + "grad_norm": 1.6590182781219482, + "learning_rate": 2.7950000000000003e-06, + "loss": 0.0689, + "step": 1764 + }, + { + "epoch": 2.088139603667554, + "grad_norm": 1.6239395141601562, + "learning_rate": 2.79375e-06, + "loss": 0.0691, + "step": 1765 + }, + { + "epoch": 2.0893226855959774, + "grad_norm": 1.4699358940124512, + "learning_rate": 2.7925000000000004e-06, + "loss": 0.0643, + "step": 1766 + }, + { + "epoch": 2.090505767524401, + "grad_norm": 1.3859870433807373, + "learning_rate": 2.79125e-06, + "loss": 0.069, + "step": 1767 + }, + { + "epoch": 2.0916888494528245, + "grad_norm": 1.6852190494537354, + "learning_rate": 2.7900000000000004e-06, + "loss": 0.0649, + "step": 1768 + }, + { + "epoch": 2.0928719313812483, + "grad_norm": 1.805801272392273, + "learning_rate": 2.78875e-06, + "loss": 0.067, + "step": 1769 + }, + { + "epoch": 2.0940550133096716, + "grad_norm": 1.6838639974594116, + "learning_rate": 2.7875000000000004e-06, + "loss": 0.0734, + "step": 1770 + }, + { + "epoch": 2.0952380952380953, + "grad_norm": 1.3731379508972168, + "learning_rate": 2.7862500000000002e-06, + "loss": 0.0665, + "step": 1771 + }, + { + "epoch": 2.0964211771665187, + "grad_norm": 1.4686869382858276, + "learning_rate": 2.7850000000000004e-06, + "loss": 0.0675, + "step": 1772 + }, + { + "epoch": 2.0976042590949424, + "grad_norm": 1.3909939527511597, + "learning_rate": 2.7837500000000002e-06, + "loss": 0.0667, + "step": 1773 + }, + { + "epoch": 2.0987873410233657, + "grad_norm": 1.5428329706192017, + "learning_rate": 2.7825e-06, + "loss": 0.0759, + "step": 1774 + }, + { + "epoch": 2.0999704229517895, + "grad_norm": 1.3300358057022095, + "learning_rate": 2.7812500000000003e-06, + "loss": 0.068, + "step": 1775 + }, + { + "epoch": 2.101153504880213, + "grad_norm": 1.4790689945220947, + "learning_rate": 2.7800000000000005e-06, + "loss": 0.0666, + "step": 1776 + }, + { + "epoch": 2.1023365868086366, + "grad_norm": 1.3302521705627441, + "learning_rate": 2.77875e-06, + "loss": 0.0705, + "step": 1777 + }, + { + "epoch": 2.10351966873706, + "grad_norm": 1.323825478553772, + "learning_rate": 2.7775e-06, + "loss": 0.064, + "step": 1778 + }, + { + "epoch": 2.1047027506654836, + "grad_norm": 1.4802190065383911, + "learning_rate": 2.7762500000000003e-06, + "loss": 0.0715, + "step": 1779 + }, + { + "epoch": 2.105885832593907, + "grad_norm": 1.6523406505584717, + "learning_rate": 2.7750000000000005e-06, + "loss": 0.0721, + "step": 1780 + }, + { + "epoch": 2.1070689145223307, + "grad_norm": 1.4791831970214844, + "learning_rate": 2.77375e-06, + "loss": 0.0722, + "step": 1781 + }, + { + "epoch": 2.108251996450754, + "grad_norm": 1.1588783264160156, + "learning_rate": 2.7725e-06, + "loss": 0.0667, + "step": 1782 + }, + { + "epoch": 2.1094350783791778, + "grad_norm": 1.9546269178390503, + "learning_rate": 2.7712500000000004e-06, + "loss": 0.0703, + "step": 1783 + }, + { + "epoch": 2.1106181603076015, + "grad_norm": 1.9815466403961182, + "learning_rate": 2.7700000000000006e-06, + "loss": 0.0728, + "step": 1784 + }, + { + "epoch": 2.111801242236025, + "grad_norm": 1.893527626991272, + "learning_rate": 2.76875e-06, + "loss": 0.0655, + "step": 1785 + }, + { + "epoch": 2.1129843241644486, + "grad_norm": 1.2941248416900635, + "learning_rate": 2.7675e-06, + "loss": 0.065, + "step": 1786 + }, + { + "epoch": 2.114167406092872, + "grad_norm": 1.4494199752807617, + "learning_rate": 2.7662500000000004e-06, + "loss": 0.07, + "step": 1787 + }, + { + "epoch": 2.1153504880212957, + "grad_norm": 2.1043918132781982, + "learning_rate": 2.7650000000000006e-06, + "loss": 0.0697, + "step": 1788 + }, + { + "epoch": 2.116533569949719, + "grad_norm": 1.7324215173721313, + "learning_rate": 2.76375e-06, + "loss": 0.08, + "step": 1789 + }, + { + "epoch": 2.1177166518781427, + "grad_norm": 1.6263593435287476, + "learning_rate": 2.7625000000000002e-06, + "loss": 0.0705, + "step": 1790 + }, + { + "epoch": 2.118899733806566, + "grad_norm": 1.5987566709518433, + "learning_rate": 2.7612500000000005e-06, + "loss": 0.0719, + "step": 1791 + }, + { + "epoch": 2.12008281573499, + "grad_norm": 1.5921862125396729, + "learning_rate": 2.7600000000000003e-06, + "loss": 0.0649, + "step": 1792 + }, + { + "epoch": 2.121265897663413, + "grad_norm": 1.612477421760559, + "learning_rate": 2.75875e-06, + "loss": 0.0655, + "step": 1793 + }, + { + "epoch": 2.122448979591837, + "grad_norm": 1.7388248443603516, + "learning_rate": 2.7575000000000003e-06, + "loss": 0.068, + "step": 1794 + }, + { + "epoch": 2.12363206152026, + "grad_norm": 1.6547331809997559, + "learning_rate": 2.7562500000000005e-06, + "loss": 0.0742, + "step": 1795 + }, + { + "epoch": 2.124815143448684, + "grad_norm": 1.6941269636154175, + "learning_rate": 2.7550000000000003e-06, + "loss": 0.0735, + "step": 1796 + }, + { + "epoch": 2.1259982253771073, + "grad_norm": 1.4354699850082397, + "learning_rate": 2.75375e-06, + "loss": 0.0691, + "step": 1797 + }, + { + "epoch": 2.127181307305531, + "grad_norm": 1.4303233623504639, + "learning_rate": 2.7525000000000003e-06, + "loss": 0.0631, + "step": 1798 + }, + { + "epoch": 2.1283643892339543, + "grad_norm": 1.6038075685501099, + "learning_rate": 2.75125e-06, + "loss": 0.0696, + "step": 1799 + }, + { + "epoch": 2.129547471162378, + "grad_norm": 1.5992631912231445, + "learning_rate": 2.7500000000000004e-06, + "loss": 0.0633, + "step": 1800 + }, + { + "epoch": 2.1307305530908014, + "grad_norm": 1.501910924911499, + "learning_rate": 2.74875e-06, + "loss": 0.0667, + "step": 1801 + }, + { + "epoch": 2.131913635019225, + "grad_norm": 1.5506956577301025, + "learning_rate": 2.7475000000000004e-06, + "loss": 0.0699, + "step": 1802 + }, + { + "epoch": 2.1330967169476485, + "grad_norm": 1.4171082973480225, + "learning_rate": 2.74625e-06, + "loss": 0.0689, + "step": 1803 + }, + { + "epoch": 2.134279798876072, + "grad_norm": 1.5881175994873047, + "learning_rate": 2.7450000000000004e-06, + "loss": 0.0666, + "step": 1804 + }, + { + "epoch": 2.1354628808044955, + "grad_norm": 1.6434916257858276, + "learning_rate": 2.74375e-06, + "loss": 0.0748, + "step": 1805 + }, + { + "epoch": 2.1366459627329193, + "grad_norm": 1.6454490423202515, + "learning_rate": 2.7425e-06, + "loss": 0.0658, + "step": 1806 + }, + { + "epoch": 2.1378290446613426, + "grad_norm": 1.7229312658309937, + "learning_rate": 2.7412500000000002e-06, + "loss": 0.0701, + "step": 1807 + }, + { + "epoch": 2.1390121265897664, + "grad_norm": 1.5081415176391602, + "learning_rate": 2.7400000000000004e-06, + "loss": 0.0741, + "step": 1808 + }, + { + "epoch": 2.1401952085181897, + "grad_norm": 1.8712142705917358, + "learning_rate": 2.73875e-06, + "loss": 0.0734, + "step": 1809 + }, + { + "epoch": 2.1413782904466134, + "grad_norm": 1.6502128839492798, + "learning_rate": 2.7375e-06, + "loss": 0.0763, + "step": 1810 + }, + { + "epoch": 2.142561372375037, + "grad_norm": 1.4047259092330933, + "learning_rate": 2.7362500000000003e-06, + "loss": 0.0753, + "step": 1811 + }, + { + "epoch": 2.1437444543034605, + "grad_norm": 1.6250574588775635, + "learning_rate": 2.7350000000000005e-06, + "loss": 0.0719, + "step": 1812 + }, + { + "epoch": 2.1449275362318843, + "grad_norm": 1.8443610668182373, + "learning_rate": 2.73375e-06, + "loss": 0.0727, + "step": 1813 + }, + { + "epoch": 2.1461106181603076, + "grad_norm": 1.656159520149231, + "learning_rate": 2.7325e-06, + "loss": 0.0731, + "step": 1814 + }, + { + "epoch": 2.1472937000887313, + "grad_norm": 1.6257644891738892, + "learning_rate": 2.7312500000000003e-06, + "loss": 0.0788, + "step": 1815 + }, + { + "epoch": 2.1484767820171546, + "grad_norm": 1.5173447132110596, + "learning_rate": 2.7300000000000005e-06, + "loss": 0.065, + "step": 1816 + }, + { + "epoch": 2.1496598639455784, + "grad_norm": 1.4590741395950317, + "learning_rate": 2.72875e-06, + "loss": 0.0771, + "step": 1817 + }, + { + "epoch": 2.1508429458740017, + "grad_norm": 1.3242361545562744, + "learning_rate": 2.7275e-06, + "loss": 0.0613, + "step": 1818 + }, + { + "epoch": 2.1520260278024255, + "grad_norm": 1.4892522096633911, + "learning_rate": 2.7262500000000004e-06, + "loss": 0.0661, + "step": 1819 + }, + { + "epoch": 2.153209109730849, + "grad_norm": 1.764828085899353, + "learning_rate": 2.7250000000000006e-06, + "loss": 0.0693, + "step": 1820 + }, + { + "epoch": 2.1543921916592725, + "grad_norm": 1.4587719440460205, + "learning_rate": 2.72375e-06, + "loss": 0.0753, + "step": 1821 + }, + { + "epoch": 2.155575273587696, + "grad_norm": 1.5698349475860596, + "learning_rate": 2.7225e-06, + "loss": 0.069, + "step": 1822 + }, + { + "epoch": 2.1567583555161196, + "grad_norm": 1.7991288900375366, + "learning_rate": 2.7212500000000004e-06, + "loss": 0.0673, + "step": 1823 + }, + { + "epoch": 2.157941437444543, + "grad_norm": 1.6338173151016235, + "learning_rate": 2.7200000000000002e-06, + "loss": 0.0678, + "step": 1824 + }, + { + "epoch": 2.1591245193729667, + "grad_norm": 1.6246289014816284, + "learning_rate": 2.71875e-06, + "loss": 0.0709, + "step": 1825 + }, + { + "epoch": 2.16030760130139, + "grad_norm": 1.5102070569992065, + "learning_rate": 2.7175000000000002e-06, + "loss": 0.0681, + "step": 1826 + }, + { + "epoch": 2.1614906832298137, + "grad_norm": 1.7507245540618896, + "learning_rate": 2.7162500000000005e-06, + "loss": 0.0674, + "step": 1827 + }, + { + "epoch": 2.162673765158237, + "grad_norm": 1.7262579202651978, + "learning_rate": 2.7150000000000003e-06, + "loss": 0.0713, + "step": 1828 + }, + { + "epoch": 2.163856847086661, + "grad_norm": 1.7287753820419312, + "learning_rate": 2.71375e-06, + "loss": 0.0697, + "step": 1829 + }, + { + "epoch": 2.165039929015084, + "grad_norm": 1.4084886312484741, + "learning_rate": 2.7125000000000003e-06, + "loss": 0.0792, + "step": 1830 + }, + { + "epoch": 2.166223010943508, + "grad_norm": 1.449167013168335, + "learning_rate": 2.71125e-06, + "loss": 0.077, + "step": 1831 + }, + { + "epoch": 2.167406092871931, + "grad_norm": 1.5783612728118896, + "learning_rate": 2.7100000000000003e-06, + "loss": 0.0691, + "step": 1832 + }, + { + "epoch": 2.168589174800355, + "grad_norm": 1.5903551578521729, + "learning_rate": 2.70875e-06, + "loss": 0.0708, + "step": 1833 + }, + { + "epoch": 2.1697722567287783, + "grad_norm": 1.3953698873519897, + "learning_rate": 2.7075000000000003e-06, + "loss": 0.0656, + "step": 1834 + }, + { + "epoch": 2.170955338657202, + "grad_norm": 1.5200599431991577, + "learning_rate": 2.70625e-06, + "loss": 0.0706, + "step": 1835 + }, + { + "epoch": 2.1721384205856253, + "grad_norm": 1.55244779586792, + "learning_rate": 2.7050000000000004e-06, + "loss": 0.0693, + "step": 1836 + }, + { + "epoch": 2.173321502514049, + "grad_norm": 1.7721471786499023, + "learning_rate": 2.70375e-06, + "loss": 0.0756, + "step": 1837 + }, + { + "epoch": 2.174504584442473, + "grad_norm": 1.7910430431365967, + "learning_rate": 2.7025e-06, + "loss": 0.0702, + "step": 1838 + }, + { + "epoch": 2.175687666370896, + "grad_norm": 1.5592271089553833, + "learning_rate": 2.70125e-06, + "loss": 0.0754, + "step": 1839 + }, + { + "epoch": 2.17687074829932, + "grad_norm": 1.461314082145691, + "learning_rate": 2.7000000000000004e-06, + "loss": 0.0739, + "step": 1840 + }, + { + "epoch": 2.1780538302277432, + "grad_norm": 1.8456028699874878, + "learning_rate": 2.6987499999999998e-06, + "loss": 0.0706, + "step": 1841 + }, + { + "epoch": 2.179236912156167, + "grad_norm": 1.4254565238952637, + "learning_rate": 2.6975e-06, + "loss": 0.0685, + "step": 1842 + }, + { + "epoch": 2.1804199940845903, + "grad_norm": 1.7270379066467285, + "learning_rate": 2.6962500000000002e-06, + "loss": 0.0721, + "step": 1843 + }, + { + "epoch": 2.181603076013014, + "grad_norm": 1.4672541618347168, + "learning_rate": 2.6950000000000005e-06, + "loss": 0.0678, + "step": 1844 + }, + { + "epoch": 2.1827861579414374, + "grad_norm": 1.6260532140731812, + "learning_rate": 2.69375e-06, + "loss": 0.0654, + "step": 1845 + }, + { + "epoch": 2.183969239869861, + "grad_norm": 1.5983856916427612, + "learning_rate": 2.6925e-06, + "loss": 0.0791, + "step": 1846 + }, + { + "epoch": 2.1851523217982844, + "grad_norm": 1.5334899425506592, + "learning_rate": 2.6912500000000003e-06, + "loss": 0.0687, + "step": 1847 + }, + { + "epoch": 2.186335403726708, + "grad_norm": 2.2006728649139404, + "learning_rate": 2.6900000000000005e-06, + "loss": 0.0748, + "step": 1848 + }, + { + "epoch": 2.1875184856551315, + "grad_norm": 1.589093565940857, + "learning_rate": 2.68875e-06, + "loss": 0.069, + "step": 1849 + }, + { + "epoch": 2.1887015675835553, + "grad_norm": 1.8567482233047485, + "learning_rate": 2.6875e-06, + "loss": 0.0766, + "step": 1850 + }, + { + "epoch": 2.1898846495119786, + "grad_norm": 1.6383507251739502, + "learning_rate": 2.6862500000000003e-06, + "loss": 0.073, + "step": 1851 + }, + { + "epoch": 2.1910677314404023, + "grad_norm": 1.3550701141357422, + "learning_rate": 2.6850000000000006e-06, + "loss": 0.066, + "step": 1852 + }, + { + "epoch": 2.1922508133688257, + "grad_norm": 1.3554604053497314, + "learning_rate": 2.68375e-06, + "loss": 0.0662, + "step": 1853 + }, + { + "epoch": 2.1934338952972494, + "grad_norm": 1.5562314987182617, + "learning_rate": 2.6825e-06, + "loss": 0.068, + "step": 1854 + }, + { + "epoch": 2.1946169772256727, + "grad_norm": 1.9308035373687744, + "learning_rate": 2.6812500000000004e-06, + "loss": 0.0735, + "step": 1855 + }, + { + "epoch": 2.1958000591540965, + "grad_norm": 1.7272809743881226, + "learning_rate": 2.68e-06, + "loss": 0.0746, + "step": 1856 + }, + { + "epoch": 2.19698314108252, + "grad_norm": 1.7436461448669434, + "learning_rate": 2.67875e-06, + "loss": 0.0719, + "step": 1857 + }, + { + "epoch": 2.1981662230109436, + "grad_norm": 1.5499225854873657, + "learning_rate": 2.6775e-06, + "loss": 0.0693, + "step": 1858 + }, + { + "epoch": 2.199349304939367, + "grad_norm": 1.4392763376235962, + "learning_rate": 2.6762500000000004e-06, + "loss": 0.0649, + "step": 1859 + }, + { + "epoch": 2.2005323868677906, + "grad_norm": 1.4923336505889893, + "learning_rate": 2.6750000000000002e-06, + "loss": 0.0685, + "step": 1860 + }, + { + "epoch": 2.201715468796214, + "grad_norm": 1.3379985094070435, + "learning_rate": 2.67375e-06, + "loss": 0.0763, + "step": 1861 + }, + { + "epoch": 2.2028985507246377, + "grad_norm": 1.6008877754211426, + "learning_rate": 2.6725000000000002e-06, + "loss": 0.0681, + "step": 1862 + }, + { + "epoch": 2.204081632653061, + "grad_norm": 1.7898168563842773, + "learning_rate": 2.67125e-06, + "loss": 0.074, + "step": 1863 + }, + { + "epoch": 2.2052647145814848, + "grad_norm": 1.7972100973129272, + "learning_rate": 2.6700000000000003e-06, + "loss": 0.0743, + "step": 1864 + }, + { + "epoch": 2.2064477965099085, + "grad_norm": 1.5675036907196045, + "learning_rate": 2.66875e-06, + "loss": 0.0722, + "step": 1865 + }, + { + "epoch": 2.207630878438332, + "grad_norm": 1.8163728713989258, + "learning_rate": 2.6675000000000003e-06, + "loss": 0.0755, + "step": 1866 + }, + { + "epoch": 2.2088139603667556, + "grad_norm": 1.8965073823928833, + "learning_rate": 2.66625e-06, + "loss": 0.0727, + "step": 1867 + }, + { + "epoch": 2.209997042295179, + "grad_norm": 1.6833105087280273, + "learning_rate": 2.6650000000000003e-06, + "loss": 0.0692, + "step": 1868 + }, + { + "epoch": 2.2111801242236027, + "grad_norm": 1.7324092388153076, + "learning_rate": 2.66375e-06, + "loss": 0.081, + "step": 1869 + }, + { + "epoch": 2.212363206152026, + "grad_norm": 1.5365663766860962, + "learning_rate": 2.6625e-06, + "loss": 0.0742, + "step": 1870 + }, + { + "epoch": 2.2135462880804497, + "grad_norm": 1.5425857305526733, + "learning_rate": 2.66125e-06, + "loss": 0.0762, + "step": 1871 + }, + { + "epoch": 2.214729370008873, + "grad_norm": 1.5223948955535889, + "learning_rate": 2.6600000000000004e-06, + "loss": 0.0729, + "step": 1872 + }, + { + "epoch": 2.215912451937297, + "grad_norm": 1.7245705127716064, + "learning_rate": 2.6587499999999997e-06, + "loss": 0.0722, + "step": 1873 + }, + { + "epoch": 2.21709553386572, + "grad_norm": 1.6245890855789185, + "learning_rate": 2.6575e-06, + "loss": 0.0692, + "step": 1874 + }, + { + "epoch": 2.218278615794144, + "grad_norm": 1.4073858261108398, + "learning_rate": 2.65625e-06, + "loss": 0.0699, + "step": 1875 + }, + { + "epoch": 2.219461697722567, + "grad_norm": 1.5762921571731567, + "learning_rate": 2.6550000000000004e-06, + "loss": 0.0753, + "step": 1876 + }, + { + "epoch": 2.220644779650991, + "grad_norm": 1.6480696201324463, + "learning_rate": 2.6537500000000006e-06, + "loss": 0.0734, + "step": 1877 + }, + { + "epoch": 2.2218278615794143, + "grad_norm": 1.4507322311401367, + "learning_rate": 2.6525e-06, + "loss": 0.0726, + "step": 1878 + }, + { + "epoch": 2.223010943507838, + "grad_norm": 1.8770027160644531, + "learning_rate": 2.6512500000000002e-06, + "loss": 0.0698, + "step": 1879 + }, + { + "epoch": 2.2241940254362613, + "grad_norm": 1.511333703994751, + "learning_rate": 2.6500000000000005e-06, + "loss": 0.0706, + "step": 1880 + }, + { + "epoch": 2.225377107364685, + "grad_norm": 1.4210904836654663, + "learning_rate": 2.6487500000000007e-06, + "loss": 0.0678, + "step": 1881 + }, + { + "epoch": 2.2265601892931084, + "grad_norm": 1.6671801805496216, + "learning_rate": 2.6475e-06, + "loss": 0.0666, + "step": 1882 + }, + { + "epoch": 2.227743271221532, + "grad_norm": 1.4927165508270264, + "learning_rate": 2.6462500000000003e-06, + "loss": 0.0659, + "step": 1883 + }, + { + "epoch": 2.2289263531499555, + "grad_norm": 1.5569570064544678, + "learning_rate": 2.6450000000000005e-06, + "loss": 0.0641, + "step": 1884 + }, + { + "epoch": 2.230109435078379, + "grad_norm": 1.6338224411010742, + "learning_rate": 2.6437500000000003e-06, + "loss": 0.0649, + "step": 1885 + }, + { + "epoch": 2.2312925170068025, + "grad_norm": 1.6820464134216309, + "learning_rate": 2.6425e-06, + "loss": 0.0665, + "step": 1886 + }, + { + "epoch": 2.2324755989352263, + "grad_norm": 2.3219687938690186, + "learning_rate": 2.6412500000000003e-06, + "loss": 0.0751, + "step": 1887 + }, + { + "epoch": 2.2336586808636496, + "grad_norm": 1.4572944641113281, + "learning_rate": 2.64e-06, + "loss": 0.0699, + "step": 1888 + }, + { + "epoch": 2.2348417627920734, + "grad_norm": 1.5677744150161743, + "learning_rate": 2.6387500000000004e-06, + "loss": 0.0752, + "step": 1889 + }, + { + "epoch": 2.2360248447204967, + "grad_norm": 1.6179250478744507, + "learning_rate": 2.6375e-06, + "loss": 0.0724, + "step": 1890 + }, + { + "epoch": 2.2372079266489204, + "grad_norm": 1.5209064483642578, + "learning_rate": 2.6362500000000004e-06, + "loss": 0.0656, + "step": 1891 + }, + { + "epoch": 2.238391008577344, + "grad_norm": 1.1926546096801758, + "learning_rate": 2.635e-06, + "loss": 0.0695, + "step": 1892 + }, + { + "epoch": 2.2395740905057675, + "grad_norm": 1.6804983615875244, + "learning_rate": 2.6337500000000004e-06, + "loss": 0.0669, + "step": 1893 + }, + { + "epoch": 2.2407571724341913, + "grad_norm": 1.524249792098999, + "learning_rate": 2.6325e-06, + "loss": 0.0664, + "step": 1894 + }, + { + "epoch": 2.2419402543626146, + "grad_norm": 1.5512033700942993, + "learning_rate": 2.63125e-06, + "loss": 0.0725, + "step": 1895 + }, + { + "epoch": 2.2431233362910383, + "grad_norm": 1.66383695602417, + "learning_rate": 2.6300000000000002e-06, + "loss": 0.0714, + "step": 1896 + }, + { + "epoch": 2.2443064182194616, + "grad_norm": 1.8067792654037476, + "learning_rate": 2.6287500000000005e-06, + "loss": 0.0688, + "step": 1897 + }, + { + "epoch": 2.2454895001478854, + "grad_norm": 1.437890648841858, + "learning_rate": 2.6275000000000003e-06, + "loss": 0.0654, + "step": 1898 + }, + { + "epoch": 2.2466725820763087, + "grad_norm": 1.6170541048049927, + "learning_rate": 2.62625e-06, + "loss": 0.0704, + "step": 1899 + }, + { + "epoch": 2.2478556640047325, + "grad_norm": 1.6636452674865723, + "learning_rate": 2.6250000000000003e-06, + "loss": 0.0711, + "step": 1900 + }, + { + "epoch": 2.249038745933156, + "grad_norm": 1.5108683109283447, + "learning_rate": 2.6237500000000005e-06, + "loss": 0.0686, + "step": 1901 + }, + { + "epoch": 2.2502218278615795, + "grad_norm": 1.5376853942871094, + "learning_rate": 2.6225e-06, + "loss": 0.069, + "step": 1902 + }, + { + "epoch": 2.251404909790003, + "grad_norm": 1.5273501873016357, + "learning_rate": 2.62125e-06, + "loss": 0.0733, + "step": 1903 + }, + { + "epoch": 2.2525879917184266, + "grad_norm": 1.7660890817642212, + "learning_rate": 2.6200000000000003e-06, + "loss": 0.0742, + "step": 1904 + }, + { + "epoch": 2.25377107364685, + "grad_norm": 1.5688343048095703, + "learning_rate": 2.6187500000000005e-06, + "loss": 0.0639, + "step": 1905 + }, + { + "epoch": 2.2549541555752737, + "grad_norm": 1.509378433227539, + "learning_rate": 2.6175e-06, + "loss": 0.0751, + "step": 1906 + }, + { + "epoch": 2.256137237503697, + "grad_norm": 1.512449860572815, + "learning_rate": 2.61625e-06, + "loss": 0.0685, + "step": 1907 + }, + { + "epoch": 2.2573203194321207, + "grad_norm": 1.287857174873352, + "learning_rate": 2.6150000000000004e-06, + "loss": 0.0667, + "step": 1908 + }, + { + "epoch": 2.258503401360544, + "grad_norm": 1.5351779460906982, + "learning_rate": 2.6137500000000006e-06, + "loss": 0.069, + "step": 1909 + }, + { + "epoch": 2.259686483288968, + "grad_norm": 1.5287269353866577, + "learning_rate": 2.6125e-06, + "loss": 0.0728, + "step": 1910 + }, + { + "epoch": 2.260869565217391, + "grad_norm": 1.5092706680297852, + "learning_rate": 2.61125e-06, + "loss": 0.0765, + "step": 1911 + }, + { + "epoch": 2.262052647145815, + "grad_norm": 1.4133044481277466, + "learning_rate": 2.6100000000000004e-06, + "loss": 0.0757, + "step": 1912 + }, + { + "epoch": 2.263235729074238, + "grad_norm": 1.6139657497406006, + "learning_rate": 2.6087500000000006e-06, + "loss": 0.0638, + "step": 1913 + }, + { + "epoch": 2.264418811002662, + "grad_norm": 1.662853479385376, + "learning_rate": 2.6075e-06, + "loss": 0.0755, + "step": 1914 + }, + { + "epoch": 2.2656018929310857, + "grad_norm": 1.5808054208755493, + "learning_rate": 2.6062500000000002e-06, + "loss": 0.0683, + "step": 1915 + }, + { + "epoch": 2.266784974859509, + "grad_norm": 1.608086347579956, + "learning_rate": 2.6050000000000005e-06, + "loss": 0.062, + "step": 1916 + }, + { + "epoch": 2.2679680567879323, + "grad_norm": 1.8371614217758179, + "learning_rate": 2.6037500000000003e-06, + "loss": 0.0723, + "step": 1917 + }, + { + "epoch": 2.269151138716356, + "grad_norm": 1.6587271690368652, + "learning_rate": 2.6025e-06, + "loss": 0.0727, + "step": 1918 + }, + { + "epoch": 2.27033422064478, + "grad_norm": 1.7423880100250244, + "learning_rate": 2.6012500000000003e-06, + "loss": 0.0763, + "step": 1919 + }, + { + "epoch": 2.271517302573203, + "grad_norm": 1.7023180723190308, + "learning_rate": 2.6e-06, + "loss": 0.0749, + "step": 1920 + }, + { + "epoch": 2.272700384501627, + "grad_norm": 1.5370951890945435, + "learning_rate": 2.5987500000000003e-06, + "loss": 0.0742, + "step": 1921 + }, + { + "epoch": 2.2738834664300502, + "grad_norm": 1.621019721031189, + "learning_rate": 2.5975e-06, + "loss": 0.0773, + "step": 1922 + }, + { + "epoch": 2.275066548358474, + "grad_norm": 1.4700099229812622, + "learning_rate": 2.5962500000000003e-06, + "loss": 0.0705, + "step": 1923 + }, + { + "epoch": 2.2762496302868973, + "grad_norm": 1.5520740747451782, + "learning_rate": 2.595e-06, + "loss": 0.0716, + "step": 1924 + }, + { + "epoch": 2.277432712215321, + "grad_norm": 1.481793761253357, + "learning_rate": 2.5937500000000004e-06, + "loss": 0.0714, + "step": 1925 + }, + { + "epoch": 2.2786157941437444, + "grad_norm": 1.598521113395691, + "learning_rate": 2.5925e-06, + "loss": 0.0695, + "step": 1926 + }, + { + "epoch": 2.279798876072168, + "grad_norm": 1.609910011291504, + "learning_rate": 2.59125e-06, + "loss": 0.0725, + "step": 1927 + }, + { + "epoch": 2.2809819580005914, + "grad_norm": 1.4362424612045288, + "learning_rate": 2.59e-06, + "loss": 0.0691, + "step": 1928 + }, + { + "epoch": 2.282165039929015, + "grad_norm": 1.5158742666244507, + "learning_rate": 2.5887500000000004e-06, + "loss": 0.0762, + "step": 1929 + }, + { + "epoch": 2.2833481218574385, + "grad_norm": 1.6810766458511353, + "learning_rate": 2.5875000000000002e-06, + "loss": 0.0734, + "step": 1930 + }, + { + "epoch": 2.2845312037858623, + "grad_norm": 1.9048405885696411, + "learning_rate": 2.58625e-06, + "loss": 0.0705, + "step": 1931 + }, + { + "epoch": 2.2857142857142856, + "grad_norm": 1.5252189636230469, + "learning_rate": 2.5850000000000002e-06, + "loss": 0.0748, + "step": 1932 + }, + { + "epoch": 2.2868973676427093, + "grad_norm": 1.41681706905365, + "learning_rate": 2.5837500000000005e-06, + "loss": 0.0701, + "step": 1933 + }, + { + "epoch": 2.2880804495711327, + "grad_norm": 1.7417865991592407, + "learning_rate": 2.5825e-06, + "loss": 0.0744, + "step": 1934 + }, + { + "epoch": 2.2892635314995564, + "grad_norm": 1.5133047103881836, + "learning_rate": 2.58125e-06, + "loss": 0.0715, + "step": 1935 + }, + { + "epoch": 2.2904466134279797, + "grad_norm": 1.6629083156585693, + "learning_rate": 2.5800000000000003e-06, + "loss": 0.0697, + "step": 1936 + }, + { + "epoch": 2.2916296953564035, + "grad_norm": 1.4478265047073364, + "learning_rate": 2.5787500000000005e-06, + "loss": 0.0718, + "step": 1937 + }, + { + "epoch": 2.292812777284827, + "grad_norm": 1.8373175859451294, + "learning_rate": 2.5775e-06, + "loss": 0.066, + "step": 1938 + }, + { + "epoch": 2.2939958592132506, + "grad_norm": 1.7680788040161133, + "learning_rate": 2.57625e-06, + "loss": 0.0761, + "step": 1939 + }, + { + "epoch": 2.295178941141674, + "grad_norm": 1.867126703262329, + "learning_rate": 2.5750000000000003e-06, + "loss": 0.073, + "step": 1940 + }, + { + "epoch": 2.2963620230700976, + "grad_norm": 1.5913681983947754, + "learning_rate": 2.5737500000000006e-06, + "loss": 0.0752, + "step": 1941 + }, + { + "epoch": 2.2975451049985214, + "grad_norm": 1.4774906635284424, + "learning_rate": 2.5725e-06, + "loss": 0.0713, + "step": 1942 + }, + { + "epoch": 2.2987281869269447, + "grad_norm": 1.2258810997009277, + "learning_rate": 2.57125e-06, + "loss": 0.0749, + "step": 1943 + }, + { + "epoch": 2.299911268855368, + "grad_norm": 1.648611307144165, + "learning_rate": 2.5700000000000004e-06, + "loss": 0.0662, + "step": 1944 + }, + { + "epoch": 2.3010943507837918, + "grad_norm": 2.131798028945923, + "learning_rate": 2.5687500000000006e-06, + "loss": 0.0763, + "step": 1945 + }, + { + "epoch": 2.3022774327122155, + "grad_norm": 1.6091140508651733, + "learning_rate": 2.5675e-06, + "loss": 0.0706, + "step": 1946 + }, + { + "epoch": 2.303460514640639, + "grad_norm": 1.9047247171401978, + "learning_rate": 2.56625e-06, + "loss": 0.0675, + "step": 1947 + }, + { + "epoch": 2.3046435965690626, + "grad_norm": 1.8237653970718384, + "learning_rate": 2.5650000000000004e-06, + "loss": 0.0794, + "step": 1948 + }, + { + "epoch": 2.305826678497486, + "grad_norm": 1.7819923162460327, + "learning_rate": 2.5637500000000002e-06, + "loss": 0.0712, + "step": 1949 + }, + { + "epoch": 2.3070097604259097, + "grad_norm": 1.571602702140808, + "learning_rate": 2.5625e-06, + "loss": 0.0694, + "step": 1950 + }, + { + "epoch": 2.308192842354333, + "grad_norm": 1.7306127548217773, + "learning_rate": 2.5612500000000003e-06, + "loss": 0.0779, + "step": 1951 + }, + { + "epoch": 2.3093759242827567, + "grad_norm": 1.6948665380477905, + "learning_rate": 2.56e-06, + "loss": 0.072, + "step": 1952 + }, + { + "epoch": 2.31055900621118, + "grad_norm": 1.6307650804519653, + "learning_rate": 2.5587500000000003e-06, + "loss": 0.0726, + "step": 1953 + }, + { + "epoch": 2.311742088139604, + "grad_norm": 1.3734707832336426, + "learning_rate": 2.5575e-06, + "loss": 0.066, + "step": 1954 + }, + { + "epoch": 2.312925170068027, + "grad_norm": 1.5334596633911133, + "learning_rate": 2.5562500000000003e-06, + "loss": 0.0695, + "step": 1955 + }, + { + "epoch": 2.314108251996451, + "grad_norm": 1.6030628681182861, + "learning_rate": 2.555e-06, + "loss": 0.0674, + "step": 1956 + }, + { + "epoch": 2.315291333924874, + "grad_norm": 1.8805547952651978, + "learning_rate": 2.5537500000000003e-06, + "loss": 0.0762, + "step": 1957 + }, + { + "epoch": 2.316474415853298, + "grad_norm": 1.8350294828414917, + "learning_rate": 2.5525e-06, + "loss": 0.069, + "step": 1958 + }, + { + "epoch": 2.3176574977817213, + "grad_norm": 1.4601608514785767, + "learning_rate": 2.55125e-06, + "loss": 0.0719, + "step": 1959 + }, + { + "epoch": 2.318840579710145, + "grad_norm": 1.861150860786438, + "learning_rate": 2.55e-06, + "loss": 0.0724, + "step": 1960 + }, + { + "epoch": 2.3200236616385683, + "grad_norm": 1.7574163675308228, + "learning_rate": 2.5487500000000004e-06, + "loss": 0.0698, + "step": 1961 + }, + { + "epoch": 2.321206743566992, + "grad_norm": 1.4861247539520264, + "learning_rate": 2.5475e-06, + "loss": 0.0723, + "step": 1962 + }, + { + "epoch": 2.3223898254954154, + "grad_norm": 1.4942668676376343, + "learning_rate": 2.54625e-06, + "loss": 0.0658, + "step": 1963 + }, + { + "epoch": 2.323572907423839, + "grad_norm": 1.6905150413513184, + "learning_rate": 2.545e-06, + "loss": 0.0689, + "step": 1964 + }, + { + "epoch": 2.3247559893522625, + "grad_norm": 1.5701631307601929, + "learning_rate": 2.5437500000000004e-06, + "loss": 0.0663, + "step": 1965 + }, + { + "epoch": 2.325939071280686, + "grad_norm": 1.4122979640960693, + "learning_rate": 2.5425e-06, + "loss": 0.073, + "step": 1966 + }, + { + "epoch": 2.3271221532091095, + "grad_norm": 1.5262281894683838, + "learning_rate": 2.54125e-06, + "loss": 0.0673, + "step": 1967 + }, + { + "epoch": 2.3283052351375333, + "grad_norm": 1.505403757095337, + "learning_rate": 2.5400000000000002e-06, + "loss": 0.0758, + "step": 1968 + }, + { + "epoch": 2.329488317065957, + "grad_norm": 1.4179326295852661, + "learning_rate": 2.5387500000000005e-06, + "loss": 0.0694, + "step": 1969 + }, + { + "epoch": 2.3306713989943804, + "grad_norm": 1.506064772605896, + "learning_rate": 2.5375e-06, + "loss": 0.0655, + "step": 1970 + }, + { + "epoch": 2.3318544809228037, + "grad_norm": 1.5334657430648804, + "learning_rate": 2.53625e-06, + "loss": 0.0713, + "step": 1971 + }, + { + "epoch": 2.3330375628512274, + "grad_norm": 1.4676501750946045, + "learning_rate": 2.5350000000000003e-06, + "loss": 0.07, + "step": 1972 + }, + { + "epoch": 2.334220644779651, + "grad_norm": 1.5776160955429077, + "learning_rate": 2.5337500000000005e-06, + "loss": 0.0727, + "step": 1973 + }, + { + "epoch": 2.3354037267080745, + "grad_norm": 1.982961654663086, + "learning_rate": 2.5325e-06, + "loss": 0.0754, + "step": 1974 + }, + { + "epoch": 2.3365868086364983, + "grad_norm": 1.51229727268219, + "learning_rate": 2.53125e-06, + "loss": 0.0701, + "step": 1975 + }, + { + "epoch": 2.3377698905649216, + "grad_norm": 1.5752853155136108, + "learning_rate": 2.5300000000000003e-06, + "loss": 0.0689, + "step": 1976 + }, + { + "epoch": 2.3389529724933453, + "grad_norm": 1.4232994318008423, + "learning_rate": 2.5287500000000006e-06, + "loss": 0.0674, + "step": 1977 + }, + { + "epoch": 2.3401360544217686, + "grad_norm": 1.9391143321990967, + "learning_rate": 2.5275e-06, + "loss": 0.0771, + "step": 1978 + }, + { + "epoch": 2.3413191363501924, + "grad_norm": 1.7219654321670532, + "learning_rate": 2.52625e-06, + "loss": 0.0642, + "step": 1979 + }, + { + "epoch": 2.3425022182786157, + "grad_norm": 1.5475144386291504, + "learning_rate": 2.5250000000000004e-06, + "loss": 0.0652, + "step": 1980 + }, + { + "epoch": 2.3436853002070395, + "grad_norm": 1.5936378240585327, + "learning_rate": 2.52375e-06, + "loss": 0.0744, + "step": 1981 + }, + { + "epoch": 2.344868382135463, + "grad_norm": 1.5697582960128784, + "learning_rate": 2.5225e-06, + "loss": 0.0653, + "step": 1982 + }, + { + "epoch": 2.3460514640638865, + "grad_norm": 1.8640893697738647, + "learning_rate": 2.5212500000000002e-06, + "loss": 0.0703, + "step": 1983 + }, + { + "epoch": 2.34723454599231, + "grad_norm": 1.7506537437438965, + "learning_rate": 2.52e-06, + "loss": 0.0674, + "step": 1984 + }, + { + "epoch": 2.3484176279207336, + "grad_norm": 1.6193732023239136, + "learning_rate": 2.5187500000000002e-06, + "loss": 0.0699, + "step": 1985 + }, + { + "epoch": 2.349600709849157, + "grad_norm": 1.759811520576477, + "learning_rate": 2.5175e-06, + "loss": 0.0756, + "step": 1986 + }, + { + "epoch": 2.3507837917775807, + "grad_norm": 1.5758216381072998, + "learning_rate": 2.5162500000000003e-06, + "loss": 0.0642, + "step": 1987 + }, + { + "epoch": 2.351966873706004, + "grad_norm": 1.744160771369934, + "learning_rate": 2.515e-06, + "loss": 0.0627, + "step": 1988 + }, + { + "epoch": 2.3531499556344277, + "grad_norm": 1.6439257860183716, + "learning_rate": 2.5137500000000003e-06, + "loss": 0.076, + "step": 1989 + }, + { + "epoch": 2.354333037562851, + "grad_norm": 1.5779918432235718, + "learning_rate": 2.5125e-06, + "loss": 0.0709, + "step": 1990 + }, + { + "epoch": 2.355516119491275, + "grad_norm": 1.4346950054168701, + "learning_rate": 2.51125e-06, + "loss": 0.0672, + "step": 1991 + }, + { + "epoch": 2.356699201419698, + "grad_norm": 1.8407448530197144, + "learning_rate": 2.51e-06, + "loss": 0.074, + "step": 1992 + }, + { + "epoch": 2.357882283348122, + "grad_norm": 1.4981447458267212, + "learning_rate": 2.5087500000000003e-06, + "loss": 0.0687, + "step": 1993 + }, + { + "epoch": 2.359065365276545, + "grad_norm": 1.4815493822097778, + "learning_rate": 2.5075e-06, + "loss": 0.0716, + "step": 1994 + }, + { + "epoch": 2.360248447204969, + "grad_norm": 1.664079189300537, + "learning_rate": 2.50625e-06, + "loss": 0.0717, + "step": 1995 + }, + { + "epoch": 2.3614315291333927, + "grad_norm": 1.8014886379241943, + "learning_rate": 2.505e-06, + "loss": 0.0717, + "step": 1996 + }, + { + "epoch": 2.362614611061816, + "grad_norm": 1.4733543395996094, + "learning_rate": 2.5037500000000004e-06, + "loss": 0.0646, + "step": 1997 + }, + { + "epoch": 2.3637976929902393, + "grad_norm": 1.7145483493804932, + "learning_rate": 2.5024999999999998e-06, + "loss": 0.0691, + "step": 1998 + }, + { + "epoch": 2.364980774918663, + "grad_norm": 1.7515465021133423, + "learning_rate": 2.50125e-06, + "loss": 0.0731, + "step": 1999 + }, + { + "epoch": 2.366163856847087, + "grad_norm": 1.7340940237045288, + "learning_rate": 2.5e-06, + "loss": 0.0761, + "step": 2000 + }, + { + "epoch": 2.36734693877551, + "grad_norm": 1.4877523183822632, + "learning_rate": 2.4987500000000004e-06, + "loss": 0.0755, + "step": 2001 + }, + { + "epoch": 2.368530020703934, + "grad_norm": 1.4686580896377563, + "learning_rate": 2.4975000000000002e-06, + "loss": 0.0647, + "step": 2002 + }, + { + "epoch": 2.3697131026323572, + "grad_norm": 1.5609357357025146, + "learning_rate": 2.4962500000000005e-06, + "loss": 0.0695, + "step": 2003 + }, + { + "epoch": 2.370896184560781, + "grad_norm": 1.4268138408660889, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.0699, + "step": 2004 + }, + { + "epoch": 2.3720792664892043, + "grad_norm": 1.5535615682601929, + "learning_rate": 2.4937500000000005e-06, + "loss": 0.0714, + "step": 2005 + }, + { + "epoch": 2.373262348417628, + "grad_norm": 1.7267721891403198, + "learning_rate": 2.4925000000000003e-06, + "loss": 0.0728, + "step": 2006 + }, + { + "epoch": 2.3744454303460514, + "grad_norm": 1.6053920984268188, + "learning_rate": 2.4912500000000005e-06, + "loss": 0.072, + "step": 2007 + }, + { + "epoch": 2.375628512274475, + "grad_norm": 1.703820824623108, + "learning_rate": 2.4900000000000003e-06, + "loss": 0.0733, + "step": 2008 + }, + { + "epoch": 2.3768115942028984, + "grad_norm": 1.4888153076171875, + "learning_rate": 2.4887500000000005e-06, + "loss": 0.0662, + "step": 2009 + }, + { + "epoch": 2.377994676131322, + "grad_norm": 1.7904468774795532, + "learning_rate": 2.4875000000000003e-06, + "loss": 0.0693, + "step": 2010 + }, + { + "epoch": 2.3791777580597455, + "grad_norm": 2.0200014114379883, + "learning_rate": 2.48625e-06, + "loss": 0.0734, + "step": 2011 + }, + { + "epoch": 2.3803608399881693, + "grad_norm": 1.903519630432129, + "learning_rate": 2.4850000000000003e-06, + "loss": 0.0753, + "step": 2012 + }, + { + "epoch": 2.3815439219165926, + "grad_norm": 1.8700484037399292, + "learning_rate": 2.48375e-06, + "loss": 0.0693, + "step": 2013 + }, + { + "epoch": 2.3827270038450163, + "grad_norm": 1.5401655435562134, + "learning_rate": 2.4825000000000004e-06, + "loss": 0.073, + "step": 2014 + }, + { + "epoch": 2.3839100857734397, + "grad_norm": 1.5700738430023193, + "learning_rate": 2.48125e-06, + "loss": 0.0637, + "step": 2015 + }, + { + "epoch": 2.3850931677018634, + "grad_norm": 1.5529141426086426, + "learning_rate": 2.4800000000000004e-06, + "loss": 0.0686, + "step": 2016 + }, + { + "epoch": 2.3862762496302867, + "grad_norm": 1.4596046209335327, + "learning_rate": 2.47875e-06, + "loss": 0.0704, + "step": 2017 + }, + { + "epoch": 2.3874593315587105, + "grad_norm": 1.9036078453063965, + "learning_rate": 2.4775e-06, + "loss": 0.0724, + "step": 2018 + }, + { + "epoch": 2.388642413487134, + "grad_norm": 1.579797625541687, + "learning_rate": 2.4762500000000002e-06, + "loss": 0.0714, + "step": 2019 + }, + { + "epoch": 2.3898254954155576, + "grad_norm": 1.6999619007110596, + "learning_rate": 2.475e-06, + "loss": 0.0704, + "step": 2020 + }, + { + "epoch": 2.391008577343981, + "grad_norm": 1.6447769403457642, + "learning_rate": 2.4737500000000002e-06, + "loss": 0.0732, + "step": 2021 + }, + { + "epoch": 2.3921916592724046, + "grad_norm": 1.4696708917617798, + "learning_rate": 2.4725e-06, + "loss": 0.0692, + "step": 2022 + }, + { + "epoch": 2.3933747412008284, + "grad_norm": 1.5083743333816528, + "learning_rate": 2.4712500000000003e-06, + "loss": 0.0643, + "step": 2023 + }, + { + "epoch": 2.3945578231292517, + "grad_norm": 1.5482319593429565, + "learning_rate": 2.47e-06, + "loss": 0.0661, + "step": 2024 + }, + { + "epoch": 2.395740905057675, + "grad_norm": 1.5995029211044312, + "learning_rate": 2.4687500000000003e-06, + "loss": 0.0695, + "step": 2025 + }, + { + "epoch": 2.3969239869860988, + "grad_norm": 1.6312288045883179, + "learning_rate": 2.4675e-06, + "loss": 0.0719, + "step": 2026 + }, + { + "epoch": 2.3981070689145225, + "grad_norm": 1.7385246753692627, + "learning_rate": 2.4662500000000003e-06, + "loss": 0.072, + "step": 2027 + }, + { + "epoch": 2.399290150842946, + "grad_norm": 1.7971831560134888, + "learning_rate": 2.465e-06, + "loss": 0.0735, + "step": 2028 + }, + { + "epoch": 2.4004732327713696, + "grad_norm": 1.7829666137695312, + "learning_rate": 2.4637500000000003e-06, + "loss": 0.0734, + "step": 2029 + }, + { + "epoch": 2.401656314699793, + "grad_norm": 1.38741135597229, + "learning_rate": 2.4625e-06, + "loss": 0.0703, + "step": 2030 + }, + { + "epoch": 2.4028393966282167, + "grad_norm": 1.62864351272583, + "learning_rate": 2.4612500000000004e-06, + "loss": 0.0717, + "step": 2031 + }, + { + "epoch": 2.40402247855664, + "grad_norm": 1.6021205186843872, + "learning_rate": 2.46e-06, + "loss": 0.0734, + "step": 2032 + }, + { + "epoch": 2.4052055604850637, + "grad_norm": 1.5213662385940552, + "learning_rate": 2.4587500000000004e-06, + "loss": 0.0713, + "step": 2033 + }, + { + "epoch": 2.406388642413487, + "grad_norm": 1.5024617910385132, + "learning_rate": 2.4575e-06, + "loss": 0.0685, + "step": 2034 + }, + { + "epoch": 2.407571724341911, + "grad_norm": 1.4359750747680664, + "learning_rate": 2.4562500000000004e-06, + "loss": 0.0666, + "step": 2035 + }, + { + "epoch": 2.408754806270334, + "grad_norm": 1.8358347415924072, + "learning_rate": 2.4550000000000002e-06, + "loss": 0.0739, + "step": 2036 + }, + { + "epoch": 2.409937888198758, + "grad_norm": 1.5541536808013916, + "learning_rate": 2.4537500000000004e-06, + "loss": 0.0658, + "step": 2037 + }, + { + "epoch": 2.411120970127181, + "grad_norm": 1.6742148399353027, + "learning_rate": 2.4525000000000002e-06, + "loss": 0.0701, + "step": 2038 + }, + { + "epoch": 2.412304052055605, + "grad_norm": 1.535934329032898, + "learning_rate": 2.4512500000000005e-06, + "loss": 0.0713, + "step": 2039 + }, + { + "epoch": 2.4134871339840283, + "grad_norm": 1.7491497993469238, + "learning_rate": 2.4500000000000003e-06, + "loss": 0.0661, + "step": 2040 + }, + { + "epoch": 2.414670215912452, + "grad_norm": 1.7162517309188843, + "learning_rate": 2.4487500000000005e-06, + "loss": 0.0734, + "step": 2041 + }, + { + "epoch": 2.4158532978408753, + "grad_norm": 1.3432501554489136, + "learning_rate": 2.4475000000000003e-06, + "loss": 0.0599, + "step": 2042 + }, + { + "epoch": 2.417036379769299, + "grad_norm": 1.5828800201416016, + "learning_rate": 2.44625e-06, + "loss": 0.0665, + "step": 2043 + }, + { + "epoch": 2.4182194616977224, + "grad_norm": 1.6799601316452026, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.0765, + "step": 2044 + }, + { + "epoch": 2.419402543626146, + "grad_norm": 1.6081180572509766, + "learning_rate": 2.44375e-06, + "loss": 0.0704, + "step": 2045 + }, + { + "epoch": 2.4205856255545695, + "grad_norm": 1.8534668684005737, + "learning_rate": 2.4425000000000003e-06, + "loss": 0.0775, + "step": 2046 + }, + { + "epoch": 2.421768707482993, + "grad_norm": 1.6720104217529297, + "learning_rate": 2.44125e-06, + "loss": 0.0705, + "step": 2047 + }, + { + "epoch": 2.4229517894114165, + "grad_norm": 1.4831020832061768, + "learning_rate": 2.4400000000000004e-06, + "loss": 0.0752, + "step": 2048 + }, + { + "epoch": 2.4241348713398403, + "grad_norm": 1.4934359788894653, + "learning_rate": 2.43875e-06, + "loss": 0.0684, + "step": 2049 + }, + { + "epoch": 2.425317953268264, + "grad_norm": 1.9425429105758667, + "learning_rate": 2.4375e-06, + "loss": 0.0723, + "step": 2050 + }, + { + "epoch": 2.4265010351966874, + "grad_norm": 1.6128356456756592, + "learning_rate": 2.43625e-06, + "loss": 0.0687, + "step": 2051 + }, + { + "epoch": 2.4276841171251107, + "grad_norm": 1.4396827220916748, + "learning_rate": 2.435e-06, + "loss": 0.068, + "step": 2052 + }, + { + "epoch": 2.4288671990535344, + "grad_norm": 1.639893889427185, + "learning_rate": 2.43375e-06, + "loss": 0.0638, + "step": 2053 + }, + { + "epoch": 2.430050280981958, + "grad_norm": 1.9025620222091675, + "learning_rate": 2.4325e-06, + "loss": 0.0689, + "step": 2054 + }, + { + "epoch": 2.4312333629103815, + "grad_norm": 1.4050815105438232, + "learning_rate": 2.4312500000000002e-06, + "loss": 0.0641, + "step": 2055 + }, + { + "epoch": 2.4324164448388053, + "grad_norm": 1.6220033168792725, + "learning_rate": 2.43e-06, + "loss": 0.0672, + "step": 2056 + }, + { + "epoch": 2.4335995267672286, + "grad_norm": 1.6734211444854736, + "learning_rate": 2.4287500000000003e-06, + "loss": 0.0707, + "step": 2057 + }, + { + "epoch": 2.4347826086956523, + "grad_norm": 1.4585671424865723, + "learning_rate": 2.4275e-06, + "loss": 0.0724, + "step": 2058 + }, + { + "epoch": 2.4359656906240756, + "grad_norm": 1.5808364152908325, + "learning_rate": 2.4262500000000003e-06, + "loss": 0.0687, + "step": 2059 + }, + { + "epoch": 2.4371487725524994, + "grad_norm": 1.6823928356170654, + "learning_rate": 2.425e-06, + "loss": 0.0736, + "step": 2060 + }, + { + "epoch": 2.4383318544809227, + "grad_norm": 1.6617430448532104, + "learning_rate": 2.4237500000000003e-06, + "loss": 0.0723, + "step": 2061 + }, + { + "epoch": 2.4395149364093465, + "grad_norm": 1.3719147443771362, + "learning_rate": 2.4225e-06, + "loss": 0.0694, + "step": 2062 + }, + { + "epoch": 2.44069801833777, + "grad_norm": 1.5750781297683716, + "learning_rate": 2.4212500000000003e-06, + "loss": 0.0747, + "step": 2063 + }, + { + "epoch": 2.4418811002661935, + "grad_norm": 1.8747748136520386, + "learning_rate": 2.42e-06, + "loss": 0.0787, + "step": 2064 + }, + { + "epoch": 2.443064182194617, + "grad_norm": 1.6921706199645996, + "learning_rate": 2.4187500000000003e-06, + "loss": 0.0794, + "step": 2065 + }, + { + "epoch": 2.4442472641230406, + "grad_norm": 1.385337471961975, + "learning_rate": 2.4175e-06, + "loss": 0.0738, + "step": 2066 + }, + { + "epoch": 2.445430346051464, + "grad_norm": 1.720412015914917, + "learning_rate": 2.4162500000000004e-06, + "loss": 0.0729, + "step": 2067 + }, + { + "epoch": 2.4466134279798877, + "grad_norm": 1.8358008861541748, + "learning_rate": 2.415e-06, + "loss": 0.072, + "step": 2068 + }, + { + "epoch": 2.447796509908311, + "grad_norm": 1.9012980461120605, + "learning_rate": 2.4137500000000004e-06, + "loss": 0.0762, + "step": 2069 + }, + { + "epoch": 2.4489795918367347, + "grad_norm": 1.5532281398773193, + "learning_rate": 2.4125e-06, + "loss": 0.0639, + "step": 2070 + }, + { + "epoch": 2.450162673765158, + "grad_norm": 1.6946734189987183, + "learning_rate": 2.4112500000000004e-06, + "loss": 0.067, + "step": 2071 + }, + { + "epoch": 2.451345755693582, + "grad_norm": 1.4511746168136597, + "learning_rate": 2.4100000000000002e-06, + "loss": 0.071, + "step": 2072 + }, + { + "epoch": 2.452528837622005, + "grad_norm": 1.554922342300415, + "learning_rate": 2.4087500000000004e-06, + "loss": 0.0725, + "step": 2073 + }, + { + "epoch": 2.453711919550429, + "grad_norm": 1.9022164344787598, + "learning_rate": 2.4075000000000002e-06, + "loss": 0.0741, + "step": 2074 + }, + { + "epoch": 2.454895001478852, + "grad_norm": 1.545162558555603, + "learning_rate": 2.40625e-06, + "loss": 0.0677, + "step": 2075 + }, + { + "epoch": 2.456078083407276, + "grad_norm": 1.7360581159591675, + "learning_rate": 2.4050000000000003e-06, + "loss": 0.0716, + "step": 2076 + }, + { + "epoch": 2.4572611653356997, + "grad_norm": 1.3728280067443848, + "learning_rate": 2.40375e-06, + "loss": 0.0682, + "step": 2077 + }, + { + "epoch": 2.458444247264123, + "grad_norm": 1.6267781257629395, + "learning_rate": 2.4025000000000003e-06, + "loss": 0.0661, + "step": 2078 + }, + { + "epoch": 2.4596273291925463, + "grad_norm": 1.7687650918960571, + "learning_rate": 2.40125e-06, + "loss": 0.0736, + "step": 2079 + }, + { + "epoch": 2.46081041112097, + "grad_norm": 1.345981240272522, + "learning_rate": 2.4000000000000003e-06, + "loss": 0.0652, + "step": 2080 + }, + { + "epoch": 2.461993493049394, + "grad_norm": 1.5290571451187134, + "learning_rate": 2.39875e-06, + "loss": 0.0697, + "step": 2081 + }, + { + "epoch": 2.463176574977817, + "grad_norm": 1.6000640392303467, + "learning_rate": 2.3975e-06, + "loss": 0.0719, + "step": 2082 + }, + { + "epoch": 2.464359656906241, + "grad_norm": 1.741998314857483, + "learning_rate": 2.39625e-06, + "loss": 0.0745, + "step": 2083 + }, + { + "epoch": 2.4655427388346642, + "grad_norm": 1.6822015047073364, + "learning_rate": 2.395e-06, + "loss": 0.064, + "step": 2084 + }, + { + "epoch": 2.466725820763088, + "grad_norm": 1.6800205707550049, + "learning_rate": 2.39375e-06, + "loss": 0.0682, + "step": 2085 + }, + { + "epoch": 2.4679089026915113, + "grad_norm": 1.3763538599014282, + "learning_rate": 2.3925e-06, + "loss": 0.0645, + "step": 2086 + }, + { + "epoch": 2.469091984619935, + "grad_norm": 1.7704856395721436, + "learning_rate": 2.39125e-06, + "loss": 0.0688, + "step": 2087 + }, + { + "epoch": 2.4702750665483584, + "grad_norm": 1.6842044591903687, + "learning_rate": 2.39e-06, + "loss": 0.0768, + "step": 2088 + }, + { + "epoch": 2.471458148476782, + "grad_norm": 2.036588191986084, + "learning_rate": 2.38875e-06, + "loss": 0.0749, + "step": 2089 + }, + { + "epoch": 2.4726412304052054, + "grad_norm": 1.4403797388076782, + "learning_rate": 2.3875e-06, + "loss": 0.0684, + "step": 2090 + }, + { + "epoch": 2.473824312333629, + "grad_norm": 1.6161786317825317, + "learning_rate": 2.3862500000000002e-06, + "loss": 0.0706, + "step": 2091 + }, + { + "epoch": 2.4750073942620525, + "grad_norm": 1.7909787893295288, + "learning_rate": 2.385e-06, + "loss": 0.0688, + "step": 2092 + }, + { + "epoch": 2.4761904761904763, + "grad_norm": 1.490040898323059, + "learning_rate": 2.3837500000000003e-06, + "loss": 0.0719, + "step": 2093 + }, + { + "epoch": 2.4773735581188996, + "grad_norm": 1.5899009704589844, + "learning_rate": 2.3825e-06, + "loss": 0.0675, + "step": 2094 + }, + { + "epoch": 2.4785566400473233, + "grad_norm": 1.6238447427749634, + "learning_rate": 2.3812500000000003e-06, + "loss": 0.0697, + "step": 2095 + }, + { + "epoch": 2.4797397219757467, + "grad_norm": 1.3793362379074097, + "learning_rate": 2.38e-06, + "loss": 0.0658, + "step": 2096 + }, + { + "epoch": 2.4809228039041704, + "grad_norm": 1.7986705303192139, + "learning_rate": 2.3787500000000003e-06, + "loss": 0.0692, + "step": 2097 + }, + { + "epoch": 2.4821058858325937, + "grad_norm": 1.628822684288025, + "learning_rate": 2.3775e-06, + "loss": 0.07, + "step": 2098 + }, + { + "epoch": 2.4832889677610175, + "grad_norm": 1.4719997644424438, + "learning_rate": 2.3762500000000003e-06, + "loss": 0.0673, + "step": 2099 + }, + { + "epoch": 2.4844720496894412, + "grad_norm": 1.4928810596466064, + "learning_rate": 2.375e-06, + "loss": 0.0699, + "step": 2100 + }, + { + "epoch": 2.4856551316178646, + "grad_norm": 2.0756077766418457, + "learning_rate": 2.3737500000000004e-06, + "loss": 0.067, + "step": 2101 + }, + { + "epoch": 2.486838213546288, + "grad_norm": 1.628829002380371, + "learning_rate": 2.3725e-06, + "loss": 0.0682, + "step": 2102 + }, + { + "epoch": 2.4880212954747116, + "grad_norm": 1.4172298908233643, + "learning_rate": 2.3712500000000004e-06, + "loss": 0.0702, + "step": 2103 + }, + { + "epoch": 2.4892043774031354, + "grad_norm": 1.5538266897201538, + "learning_rate": 2.37e-06, + "loss": 0.075, + "step": 2104 + }, + { + "epoch": 2.4903874593315587, + "grad_norm": 1.4143210649490356, + "learning_rate": 2.3687500000000004e-06, + "loss": 0.0624, + "step": 2105 + }, + { + "epoch": 2.491570541259982, + "grad_norm": 1.8111255168914795, + "learning_rate": 2.3675e-06, + "loss": 0.0787, + "step": 2106 + }, + { + "epoch": 2.4927536231884058, + "grad_norm": 1.3706320524215698, + "learning_rate": 2.36625e-06, + "loss": 0.0665, + "step": 2107 + }, + { + "epoch": 2.4939367051168295, + "grad_norm": 1.7858434915542603, + "learning_rate": 2.3650000000000002e-06, + "loss": 0.0675, + "step": 2108 + }, + { + "epoch": 2.495119787045253, + "grad_norm": 1.665281057357788, + "learning_rate": 2.36375e-06, + "loss": 0.0745, + "step": 2109 + }, + { + "epoch": 2.4963028689736766, + "grad_norm": 1.7344304323196411, + "learning_rate": 2.3625000000000003e-06, + "loss": 0.0723, + "step": 2110 + }, + { + "epoch": 2.4974859509021, + "grad_norm": 1.6520979404449463, + "learning_rate": 2.36125e-06, + "loss": 0.0707, + "step": 2111 + }, + { + "epoch": 2.4986690328305237, + "grad_norm": 1.4923932552337646, + "learning_rate": 2.3600000000000003e-06, + "loss": 0.0722, + "step": 2112 + }, + { + "epoch": 2.499852114758947, + "grad_norm": 1.647995114326477, + "learning_rate": 2.35875e-06, + "loss": 0.0686, + "step": 2113 + }, + { + "epoch": 2.5010351966873707, + "grad_norm": 1.5371850728988647, + "learning_rate": 2.3575e-06, + "loss": 0.0724, + "step": 2114 + }, + { + "epoch": 2.502218278615794, + "grad_norm": 1.7119650840759277, + "learning_rate": 2.35625e-06, + "loss": 0.0719, + "step": 2115 + }, + { + "epoch": 2.503401360544218, + "grad_norm": 1.4010424613952637, + "learning_rate": 2.355e-06, + "loss": 0.0686, + "step": 2116 + }, + { + "epoch": 2.504584442472641, + "grad_norm": 1.3312026262283325, + "learning_rate": 2.35375e-06, + "loss": 0.0669, + "step": 2117 + }, + { + "epoch": 2.505767524401065, + "grad_norm": 1.747403621673584, + "learning_rate": 2.3525e-06, + "loss": 0.0779, + "step": 2118 + }, + { + "epoch": 2.506950606329488, + "grad_norm": 1.777378797531128, + "learning_rate": 2.35125e-06, + "loss": 0.0742, + "step": 2119 + }, + { + "epoch": 2.508133688257912, + "grad_norm": 1.6192094087600708, + "learning_rate": 2.35e-06, + "loss": 0.0683, + "step": 2120 + }, + { + "epoch": 2.5093167701863353, + "grad_norm": 1.4519423246383667, + "learning_rate": 2.34875e-06, + "loss": 0.0665, + "step": 2121 + }, + { + "epoch": 2.510499852114759, + "grad_norm": 1.6812961101531982, + "learning_rate": 2.3475e-06, + "loss": 0.0704, + "step": 2122 + }, + { + "epoch": 2.5116829340431823, + "grad_norm": 1.4318269491195679, + "learning_rate": 2.34625e-06, + "loss": 0.072, + "step": 2123 + }, + { + "epoch": 2.512866015971606, + "grad_norm": 1.6497302055358887, + "learning_rate": 2.345e-06, + "loss": 0.0688, + "step": 2124 + }, + { + "epoch": 2.5140490979000294, + "grad_norm": 1.589307188987732, + "learning_rate": 2.3437500000000002e-06, + "loss": 0.0651, + "step": 2125 + }, + { + "epoch": 2.515232179828453, + "grad_norm": 1.3222081661224365, + "learning_rate": 2.3425000000000004e-06, + "loss": 0.0677, + "step": 2126 + }, + { + "epoch": 2.516415261756877, + "grad_norm": 1.4803348779678345, + "learning_rate": 2.3412500000000002e-06, + "loss": 0.0655, + "step": 2127 + }, + { + "epoch": 2.5175983436853, + "grad_norm": 1.3884329795837402, + "learning_rate": 2.3400000000000005e-06, + "loss": 0.0648, + "step": 2128 + }, + { + "epoch": 2.5187814256137235, + "grad_norm": 1.804556965827942, + "learning_rate": 2.3387500000000003e-06, + "loss": 0.0739, + "step": 2129 + }, + { + "epoch": 2.5199645075421473, + "grad_norm": 1.784862995147705, + "learning_rate": 2.3375000000000005e-06, + "loss": 0.0748, + "step": 2130 + }, + { + "epoch": 2.521147589470571, + "grad_norm": 1.9003705978393555, + "learning_rate": 2.3362500000000003e-06, + "loss": 0.0753, + "step": 2131 + }, + { + "epoch": 2.5223306713989944, + "grad_norm": 1.9724751710891724, + "learning_rate": 2.3350000000000005e-06, + "loss": 0.0673, + "step": 2132 + }, + { + "epoch": 2.5235137533274177, + "grad_norm": 1.6086491346359253, + "learning_rate": 2.3337500000000003e-06, + "loss": 0.0674, + "step": 2133 + }, + { + "epoch": 2.5246968352558414, + "grad_norm": 1.5649945735931396, + "learning_rate": 2.3325e-06, + "loss": 0.0679, + "step": 2134 + }, + { + "epoch": 2.525879917184265, + "grad_norm": 1.6782591342926025, + "learning_rate": 2.3312500000000003e-06, + "loss": 0.0721, + "step": 2135 + }, + { + "epoch": 2.5270629991126885, + "grad_norm": 1.7564033269882202, + "learning_rate": 2.33e-06, + "loss": 0.0711, + "step": 2136 + }, + { + "epoch": 2.5282460810411123, + "grad_norm": 1.6267682313919067, + "learning_rate": 2.3287500000000004e-06, + "loss": 0.0756, + "step": 2137 + }, + { + "epoch": 2.5294291629695356, + "grad_norm": 1.7511285543441772, + "learning_rate": 2.3275e-06, + "loss": 0.07, + "step": 2138 + }, + { + "epoch": 2.5306122448979593, + "grad_norm": 1.652316689491272, + "learning_rate": 2.3262500000000004e-06, + "loss": 0.0712, + "step": 2139 + }, + { + "epoch": 2.5317953268263826, + "grad_norm": 1.9661147594451904, + "learning_rate": 2.325e-06, + "loss": 0.0703, + "step": 2140 + }, + { + "epoch": 2.5329784087548064, + "grad_norm": 1.824886679649353, + "learning_rate": 2.32375e-06, + "loss": 0.0776, + "step": 2141 + }, + { + "epoch": 2.5341614906832297, + "grad_norm": 1.6645827293395996, + "learning_rate": 2.3225e-06, + "loss": 0.0699, + "step": 2142 + }, + { + "epoch": 2.5353445726116535, + "grad_norm": 1.6906543970108032, + "learning_rate": 2.32125e-06, + "loss": 0.0681, + "step": 2143 + }, + { + "epoch": 2.536527654540077, + "grad_norm": 1.7196813821792603, + "learning_rate": 2.3200000000000002e-06, + "loss": 0.0631, + "step": 2144 + }, + { + "epoch": 2.5377107364685005, + "grad_norm": 2.102699041366577, + "learning_rate": 2.31875e-06, + "loss": 0.0735, + "step": 2145 + }, + { + "epoch": 2.538893818396924, + "grad_norm": 2.0785980224609375, + "learning_rate": 2.3175000000000003e-06, + "loss": 0.069, + "step": 2146 + }, + { + "epoch": 2.5400769003253476, + "grad_norm": 1.4073070287704468, + "learning_rate": 2.31625e-06, + "loss": 0.0688, + "step": 2147 + }, + { + "epoch": 2.541259982253771, + "grad_norm": 1.4928133487701416, + "learning_rate": 2.3150000000000003e-06, + "loss": 0.0742, + "step": 2148 + }, + { + "epoch": 2.5424430641821947, + "grad_norm": 1.7863054275512695, + "learning_rate": 2.31375e-06, + "loss": 0.0689, + "step": 2149 + }, + { + "epoch": 2.5436261461106184, + "grad_norm": 1.6399742364883423, + "learning_rate": 2.3125000000000003e-06, + "loss": 0.0677, + "step": 2150 + }, + { + "epoch": 2.5448092280390417, + "grad_norm": 1.646720051765442, + "learning_rate": 2.31125e-06, + "loss": 0.0724, + "step": 2151 + }, + { + "epoch": 2.545992309967465, + "grad_norm": 1.6679373979568481, + "learning_rate": 2.3100000000000003e-06, + "loss": 0.068, + "step": 2152 + }, + { + "epoch": 2.547175391895889, + "grad_norm": 1.3445487022399902, + "learning_rate": 2.30875e-06, + "loss": 0.0611, + "step": 2153 + }, + { + "epoch": 2.5483584738243126, + "grad_norm": 1.5276535749435425, + "learning_rate": 2.3075000000000004e-06, + "loss": 0.0687, + "step": 2154 + }, + { + "epoch": 2.549541555752736, + "grad_norm": 1.9917709827423096, + "learning_rate": 2.30625e-06, + "loss": 0.0713, + "step": 2155 + }, + { + "epoch": 2.550724637681159, + "grad_norm": 1.6489973068237305, + "learning_rate": 2.3050000000000004e-06, + "loss": 0.0665, + "step": 2156 + }, + { + "epoch": 2.551907719609583, + "grad_norm": 1.637893795967102, + "learning_rate": 2.30375e-06, + "loss": 0.0663, + "step": 2157 + }, + { + "epoch": 2.5530908015380067, + "grad_norm": 1.7272385358810425, + "learning_rate": 2.3025000000000004e-06, + "loss": 0.0698, + "step": 2158 + }, + { + "epoch": 2.55427388346643, + "grad_norm": 1.6955816745758057, + "learning_rate": 2.30125e-06, + "loss": 0.0748, + "step": 2159 + }, + { + "epoch": 2.5554569653948533, + "grad_norm": 1.4607311487197876, + "learning_rate": 2.3000000000000004e-06, + "loss": 0.0717, + "step": 2160 + }, + { + "epoch": 2.556640047323277, + "grad_norm": 1.608957290649414, + "learning_rate": 2.2987500000000002e-06, + "loss": 0.0709, + "step": 2161 + }, + { + "epoch": 2.557823129251701, + "grad_norm": 1.569519281387329, + "learning_rate": 2.2975000000000004e-06, + "loss": 0.0683, + "step": 2162 + }, + { + "epoch": 2.559006211180124, + "grad_norm": 1.5686373710632324, + "learning_rate": 2.2962500000000002e-06, + "loss": 0.0686, + "step": 2163 + }, + { + "epoch": 2.560189293108548, + "grad_norm": 1.7191416025161743, + "learning_rate": 2.2950000000000005e-06, + "loss": 0.0753, + "step": 2164 + }, + { + "epoch": 2.5613723750369712, + "grad_norm": 1.721145510673523, + "learning_rate": 2.2937500000000003e-06, + "loss": 0.0729, + "step": 2165 + }, + { + "epoch": 2.562555456965395, + "grad_norm": 1.3868467807769775, + "learning_rate": 2.2925e-06, + "loss": 0.0694, + "step": 2166 + }, + { + "epoch": 2.5637385388938183, + "grad_norm": 1.4115382432937622, + "learning_rate": 2.2912500000000003e-06, + "loss": 0.058, + "step": 2167 + }, + { + "epoch": 2.564921620822242, + "grad_norm": 1.7655788660049438, + "learning_rate": 2.29e-06, + "loss": 0.0694, + "step": 2168 + }, + { + "epoch": 2.5661047027506654, + "grad_norm": 1.714644193649292, + "learning_rate": 2.2887500000000003e-06, + "loss": 0.0668, + "step": 2169 + }, + { + "epoch": 2.567287784679089, + "grad_norm": 1.4191397428512573, + "learning_rate": 2.2875e-06, + "loss": 0.0781, + "step": 2170 + }, + { + "epoch": 2.5684708666075124, + "grad_norm": 1.9382989406585693, + "learning_rate": 2.2862500000000003e-06, + "loss": 0.068, + "step": 2171 + }, + { + "epoch": 2.569653948535936, + "grad_norm": 1.5142606496810913, + "learning_rate": 2.285e-06, + "loss": 0.0807, + "step": 2172 + }, + { + "epoch": 2.5708370304643595, + "grad_norm": 1.7104367017745972, + "learning_rate": 2.28375e-06, + "loss": 0.0738, + "step": 2173 + }, + { + "epoch": 2.5720201123927833, + "grad_norm": 1.5563592910766602, + "learning_rate": 2.2825e-06, + "loss": 0.0693, + "step": 2174 + }, + { + "epoch": 2.5732031943212066, + "grad_norm": 1.4756860733032227, + "learning_rate": 2.28125e-06, + "loss": 0.0698, + "step": 2175 + }, + { + "epoch": 2.5743862762496303, + "grad_norm": 2.0457379817962646, + "learning_rate": 2.28e-06, + "loss": 0.0871, + "step": 2176 + }, + { + "epoch": 2.575569358178054, + "grad_norm": 1.5311439037322998, + "learning_rate": 2.27875e-06, + "loss": 0.0751, + "step": 2177 + }, + { + "epoch": 2.5767524401064774, + "grad_norm": 1.643684983253479, + "learning_rate": 2.2775000000000002e-06, + "loss": 0.0654, + "step": 2178 + }, + { + "epoch": 2.5779355220349007, + "grad_norm": 1.5322777032852173, + "learning_rate": 2.27625e-06, + "loss": 0.0717, + "step": 2179 + }, + { + "epoch": 2.5791186039633245, + "grad_norm": 1.590064525604248, + "learning_rate": 2.2750000000000002e-06, + "loss": 0.0713, + "step": 2180 + }, + { + "epoch": 2.5803016858917482, + "grad_norm": 1.4851150512695312, + "learning_rate": 2.27375e-06, + "loss": 0.0691, + "step": 2181 + }, + { + "epoch": 2.5814847678201716, + "grad_norm": 1.337254285812378, + "learning_rate": 2.2725000000000003e-06, + "loss": 0.0728, + "step": 2182 + }, + { + "epoch": 2.582667849748595, + "grad_norm": 1.3873859643936157, + "learning_rate": 2.27125e-06, + "loss": 0.0721, + "step": 2183 + }, + { + "epoch": 2.5838509316770186, + "grad_norm": 1.7759919166564941, + "learning_rate": 2.2700000000000003e-06, + "loss": 0.0774, + "step": 2184 + }, + { + "epoch": 2.5850340136054424, + "grad_norm": 1.9087167978286743, + "learning_rate": 2.26875e-06, + "loss": 0.0721, + "step": 2185 + }, + { + "epoch": 2.5862170955338657, + "grad_norm": 1.447440505027771, + "learning_rate": 2.2675000000000003e-06, + "loss": 0.0748, + "step": 2186 + }, + { + "epoch": 2.587400177462289, + "grad_norm": 1.6807732582092285, + "learning_rate": 2.26625e-06, + "loss": 0.0706, + "step": 2187 + }, + { + "epoch": 2.5885832593907128, + "grad_norm": 1.769028902053833, + "learning_rate": 2.2650000000000003e-06, + "loss": 0.0791, + "step": 2188 + }, + { + "epoch": 2.5897663413191365, + "grad_norm": 1.481701374053955, + "learning_rate": 2.26375e-06, + "loss": 0.0653, + "step": 2189 + }, + { + "epoch": 2.59094942324756, + "grad_norm": 1.5643932819366455, + "learning_rate": 2.2625000000000004e-06, + "loss": 0.0759, + "step": 2190 + }, + { + "epoch": 2.5921325051759836, + "grad_norm": 1.6061702966690063, + "learning_rate": 2.26125e-06, + "loss": 0.0641, + "step": 2191 + }, + { + "epoch": 2.593315587104407, + "grad_norm": 1.6960684061050415, + "learning_rate": 2.2600000000000004e-06, + "loss": 0.0718, + "step": 2192 + }, + { + "epoch": 2.5944986690328307, + "grad_norm": 1.9704458713531494, + "learning_rate": 2.25875e-06, + "loss": 0.076, + "step": 2193 + }, + { + "epoch": 2.595681750961254, + "grad_norm": 1.7238534688949585, + "learning_rate": 2.2575000000000004e-06, + "loss": 0.0702, + "step": 2194 + }, + { + "epoch": 2.5968648328896777, + "grad_norm": 1.772412896156311, + "learning_rate": 2.25625e-06, + "loss": 0.0724, + "step": 2195 + }, + { + "epoch": 2.598047914818101, + "grad_norm": 1.5494329929351807, + "learning_rate": 2.2550000000000004e-06, + "loss": 0.0744, + "step": 2196 + }, + { + "epoch": 2.599230996746525, + "grad_norm": 1.716202974319458, + "learning_rate": 2.2537500000000002e-06, + "loss": 0.0696, + "step": 2197 + }, + { + "epoch": 2.600414078674948, + "grad_norm": 1.5587215423583984, + "learning_rate": 2.2525e-06, + "loss": 0.0648, + "step": 2198 + }, + { + "epoch": 2.601597160603372, + "grad_norm": 1.684779167175293, + "learning_rate": 2.2512500000000003e-06, + "loss": 0.0676, + "step": 2199 + }, + { + "epoch": 2.602780242531795, + "grad_norm": 1.5642940998077393, + "learning_rate": 2.25e-06, + "loss": 0.0671, + "step": 2200 + }, + { + "epoch": 2.603963324460219, + "grad_norm": 1.4968513250350952, + "learning_rate": 2.2487500000000003e-06, + "loss": 0.0702, + "step": 2201 + }, + { + "epoch": 2.6051464063886423, + "grad_norm": 1.4695079326629639, + "learning_rate": 2.2475e-06, + "loss": 0.069, + "step": 2202 + }, + { + "epoch": 2.606329488317066, + "grad_norm": 1.60423743724823, + "learning_rate": 2.2462500000000003e-06, + "loss": 0.0691, + "step": 2203 + }, + { + "epoch": 2.6075125702454898, + "grad_norm": 2.0133116245269775, + "learning_rate": 2.245e-06, + "loss": 0.0725, + "step": 2204 + }, + { + "epoch": 2.608695652173913, + "grad_norm": 1.6549969911575317, + "learning_rate": 2.24375e-06, + "loss": 0.0665, + "step": 2205 + }, + { + "epoch": 2.6098787341023364, + "grad_norm": 1.6202393770217896, + "learning_rate": 2.2425e-06, + "loss": 0.0669, + "step": 2206 + }, + { + "epoch": 2.61106181603076, + "grad_norm": 1.8169716596603394, + "learning_rate": 2.24125e-06, + "loss": 0.0715, + "step": 2207 + }, + { + "epoch": 2.612244897959184, + "grad_norm": 1.683886170387268, + "learning_rate": 2.24e-06, + "loss": 0.0666, + "step": 2208 + }, + { + "epoch": 2.6134279798876072, + "grad_norm": 1.8673646450042725, + "learning_rate": 2.23875e-06, + "loss": 0.0724, + "step": 2209 + }, + { + "epoch": 2.6146110618160305, + "grad_norm": 1.4431850910186768, + "learning_rate": 2.2375e-06, + "loss": 0.0697, + "step": 2210 + }, + { + "epoch": 2.6157941437444543, + "grad_norm": 1.5438834428787231, + "learning_rate": 2.23625e-06, + "loss": 0.0673, + "step": 2211 + }, + { + "epoch": 2.616977225672878, + "grad_norm": 1.8279645442962646, + "learning_rate": 2.235e-06, + "loss": 0.0612, + "step": 2212 + }, + { + "epoch": 2.6181603076013014, + "grad_norm": 1.7234658002853394, + "learning_rate": 2.23375e-06, + "loss": 0.069, + "step": 2213 + }, + { + "epoch": 2.6193433895297247, + "grad_norm": 1.7701345682144165, + "learning_rate": 2.2325000000000002e-06, + "loss": 0.0776, + "step": 2214 + }, + { + "epoch": 2.6205264714581484, + "grad_norm": 1.6712673902511597, + "learning_rate": 2.23125e-06, + "loss": 0.0688, + "step": 2215 + }, + { + "epoch": 2.621709553386572, + "grad_norm": 1.7470591068267822, + "learning_rate": 2.2300000000000002e-06, + "loss": 0.0723, + "step": 2216 + }, + { + "epoch": 2.6228926353149955, + "grad_norm": 1.3492313623428345, + "learning_rate": 2.22875e-06, + "loss": 0.0675, + "step": 2217 + }, + { + "epoch": 2.6240757172434193, + "grad_norm": 1.6999480724334717, + "learning_rate": 2.2275000000000003e-06, + "loss": 0.0734, + "step": 2218 + }, + { + "epoch": 2.6252587991718426, + "grad_norm": 1.2839548587799072, + "learning_rate": 2.22625e-06, + "loss": 0.0707, + "step": 2219 + }, + { + "epoch": 2.6264418811002663, + "grad_norm": 1.6676214933395386, + "learning_rate": 2.2250000000000003e-06, + "loss": 0.0665, + "step": 2220 + }, + { + "epoch": 2.6276249630286896, + "grad_norm": 1.694377064704895, + "learning_rate": 2.22375e-06, + "loss": 0.069, + "step": 2221 + }, + { + "epoch": 2.6288080449571134, + "grad_norm": 1.6888079643249512, + "learning_rate": 2.2225000000000003e-06, + "loss": 0.0728, + "step": 2222 + }, + { + "epoch": 2.6299911268855367, + "grad_norm": 1.5604629516601562, + "learning_rate": 2.22125e-06, + "loss": 0.0673, + "step": 2223 + }, + { + "epoch": 2.6311742088139605, + "grad_norm": 1.6127493381500244, + "learning_rate": 2.2200000000000003e-06, + "loss": 0.0672, + "step": 2224 + }, + { + "epoch": 2.632357290742384, + "grad_norm": 1.40291428565979, + "learning_rate": 2.21875e-06, + "loss": 0.0649, + "step": 2225 + }, + { + "epoch": 2.6335403726708075, + "grad_norm": 1.7295169830322266, + "learning_rate": 2.2175000000000004e-06, + "loss": 0.0726, + "step": 2226 + }, + { + "epoch": 2.634723454599231, + "grad_norm": 1.7472020387649536, + "learning_rate": 2.21625e-06, + "loss": 0.0717, + "step": 2227 + }, + { + "epoch": 2.6359065365276546, + "grad_norm": 1.6681567430496216, + "learning_rate": 2.2150000000000004e-06, + "loss": 0.0678, + "step": 2228 + }, + { + "epoch": 2.637089618456078, + "grad_norm": 1.8581095933914185, + "learning_rate": 2.21375e-06, + "loss": 0.0719, + "step": 2229 + }, + { + "epoch": 2.6382727003845017, + "grad_norm": 1.5567489862442017, + "learning_rate": 2.2125e-06, + "loss": 0.0679, + "step": 2230 + }, + { + "epoch": 2.6394557823129254, + "grad_norm": 1.6580990552902222, + "learning_rate": 2.2112500000000002e-06, + "loss": 0.0733, + "step": 2231 + }, + { + "epoch": 2.6406388642413487, + "grad_norm": 1.939288854598999, + "learning_rate": 2.21e-06, + "loss": 0.0763, + "step": 2232 + }, + { + "epoch": 2.641821946169772, + "grad_norm": 1.4174250364303589, + "learning_rate": 2.2087500000000002e-06, + "loss": 0.0645, + "step": 2233 + }, + { + "epoch": 2.643005028098196, + "grad_norm": 1.6383723020553589, + "learning_rate": 2.2075e-06, + "loss": 0.0711, + "step": 2234 + }, + { + "epoch": 2.6441881100266196, + "grad_norm": 1.7185078859329224, + "learning_rate": 2.2062500000000003e-06, + "loss": 0.0705, + "step": 2235 + }, + { + "epoch": 2.645371191955043, + "grad_norm": 1.4885845184326172, + "learning_rate": 2.205e-06, + "loss": 0.068, + "step": 2236 + }, + { + "epoch": 2.646554273883466, + "grad_norm": 1.4830344915390015, + "learning_rate": 2.20375e-06, + "loss": 0.0727, + "step": 2237 + }, + { + "epoch": 2.64773735581189, + "grad_norm": 1.7282239198684692, + "learning_rate": 2.2025e-06, + "loss": 0.0757, + "step": 2238 + }, + { + "epoch": 2.6489204377403137, + "grad_norm": 1.7158857583999634, + "learning_rate": 2.20125e-06, + "loss": 0.0725, + "step": 2239 + }, + { + "epoch": 2.650103519668737, + "grad_norm": 1.5162285566329956, + "learning_rate": 2.2e-06, + "loss": 0.0736, + "step": 2240 + }, + { + "epoch": 2.6512866015971603, + "grad_norm": 1.7357571125030518, + "learning_rate": 2.19875e-06, + "loss": 0.069, + "step": 2241 + }, + { + "epoch": 2.652469683525584, + "grad_norm": 1.5058177709579468, + "learning_rate": 2.1975e-06, + "loss": 0.0649, + "step": 2242 + }, + { + "epoch": 2.653652765454008, + "grad_norm": 1.7472163438796997, + "learning_rate": 2.19625e-06, + "loss": 0.0732, + "step": 2243 + }, + { + "epoch": 2.654835847382431, + "grad_norm": 1.6266549825668335, + "learning_rate": 2.195e-06, + "loss": 0.0659, + "step": 2244 + }, + { + "epoch": 2.656018929310855, + "grad_norm": 1.7285590171813965, + "learning_rate": 2.19375e-06, + "loss": 0.0652, + "step": 2245 + }, + { + "epoch": 2.6572020112392782, + "grad_norm": 1.4681788682937622, + "learning_rate": 2.1925e-06, + "loss": 0.0627, + "step": 2246 + }, + { + "epoch": 2.658385093167702, + "grad_norm": 1.4231594800949097, + "learning_rate": 2.19125e-06, + "loss": 0.0684, + "step": 2247 + }, + { + "epoch": 2.6595681750961253, + "grad_norm": 1.4145891666412354, + "learning_rate": 2.19e-06, + "loss": 0.0628, + "step": 2248 + }, + { + "epoch": 2.660751257024549, + "grad_norm": 1.6459437608718872, + "learning_rate": 2.18875e-06, + "loss": 0.0693, + "step": 2249 + }, + { + "epoch": 2.6619343389529724, + "grad_norm": 1.8473103046417236, + "learning_rate": 2.1875000000000002e-06, + "loss": 0.0665, + "step": 2250 + }, + { + "epoch": 2.663117420881396, + "grad_norm": 1.3915003538131714, + "learning_rate": 2.1862500000000005e-06, + "loss": 0.067, + "step": 2251 + }, + { + "epoch": 2.6643005028098194, + "grad_norm": 1.460813045501709, + "learning_rate": 2.1850000000000003e-06, + "loss": 0.0699, + "step": 2252 + }, + { + "epoch": 2.665483584738243, + "grad_norm": 2.0912516117095947, + "learning_rate": 2.1837500000000005e-06, + "loss": 0.074, + "step": 2253 + }, + { + "epoch": 2.6666666666666665, + "grad_norm": 1.4902088642120361, + "learning_rate": 2.1825000000000003e-06, + "loss": 0.0682, + "step": 2254 + }, + { + "epoch": 2.6678497485950903, + "grad_norm": 1.440617322921753, + "learning_rate": 2.1812500000000005e-06, + "loss": 0.0679, + "step": 2255 + }, + { + "epoch": 2.6690328305235136, + "grad_norm": 1.632251501083374, + "learning_rate": 2.1800000000000003e-06, + "loss": 0.0688, + "step": 2256 + }, + { + "epoch": 2.6702159124519373, + "grad_norm": 1.4313255548477173, + "learning_rate": 2.17875e-06, + "loss": 0.0675, + "step": 2257 + }, + { + "epoch": 2.671398994380361, + "grad_norm": 1.469079613685608, + "learning_rate": 2.1775000000000003e-06, + "loss": 0.0761, + "step": 2258 + }, + { + "epoch": 2.6725820763087844, + "grad_norm": 1.650695562362671, + "learning_rate": 2.17625e-06, + "loss": 0.0703, + "step": 2259 + }, + { + "epoch": 2.6737651582372077, + "grad_norm": 1.5478646755218506, + "learning_rate": 2.1750000000000004e-06, + "loss": 0.0698, + "step": 2260 + }, + { + "epoch": 2.6749482401656315, + "grad_norm": 1.4293920993804932, + "learning_rate": 2.17375e-06, + "loss": 0.0662, + "step": 2261 + }, + { + "epoch": 2.6761313220940552, + "grad_norm": 1.564490795135498, + "learning_rate": 2.1725000000000004e-06, + "loss": 0.0632, + "step": 2262 + }, + { + "epoch": 2.6773144040224786, + "grad_norm": 1.6599029302597046, + "learning_rate": 2.17125e-06, + "loss": 0.0692, + "step": 2263 + }, + { + "epoch": 2.678497485950902, + "grad_norm": 1.4264119863510132, + "learning_rate": 2.17e-06, + "loss": 0.062, + "step": 2264 + }, + { + "epoch": 2.6796805678793256, + "grad_norm": 1.6230655908584595, + "learning_rate": 2.16875e-06, + "loss": 0.0668, + "step": 2265 + }, + { + "epoch": 2.6808636498077494, + "grad_norm": 1.353962779045105, + "learning_rate": 2.1675e-06, + "loss": 0.0673, + "step": 2266 + }, + { + "epoch": 2.6820467317361727, + "grad_norm": 1.487084984779358, + "learning_rate": 2.1662500000000002e-06, + "loss": 0.0709, + "step": 2267 + }, + { + "epoch": 2.683229813664596, + "grad_norm": 1.8371187448501587, + "learning_rate": 2.165e-06, + "loss": 0.0698, + "step": 2268 + }, + { + "epoch": 2.6844128955930198, + "grad_norm": 1.8341972827911377, + "learning_rate": 2.1637500000000002e-06, + "loss": 0.0801, + "step": 2269 + }, + { + "epoch": 2.6855959775214435, + "grad_norm": 1.7118314504623413, + "learning_rate": 2.1625e-06, + "loss": 0.0711, + "step": 2270 + }, + { + "epoch": 2.686779059449867, + "grad_norm": 1.768996238708496, + "learning_rate": 2.1612500000000003e-06, + "loss": 0.0748, + "step": 2271 + }, + { + "epoch": 2.6879621413782906, + "grad_norm": 1.611024260520935, + "learning_rate": 2.16e-06, + "loss": 0.0649, + "step": 2272 + }, + { + "epoch": 2.689145223306714, + "grad_norm": 1.6264145374298096, + "learning_rate": 2.1587500000000003e-06, + "loss": 0.0716, + "step": 2273 + }, + { + "epoch": 2.6903283052351377, + "grad_norm": 1.5499999523162842, + "learning_rate": 2.1575e-06, + "loss": 0.0683, + "step": 2274 + }, + { + "epoch": 2.691511387163561, + "grad_norm": 1.6868284940719604, + "learning_rate": 2.1562500000000003e-06, + "loss": 0.0741, + "step": 2275 + }, + { + "epoch": 2.6926944690919847, + "grad_norm": 1.8934708833694458, + "learning_rate": 2.155e-06, + "loss": 0.0675, + "step": 2276 + }, + { + "epoch": 2.693877551020408, + "grad_norm": 1.6334190368652344, + "learning_rate": 2.1537500000000003e-06, + "loss": 0.0746, + "step": 2277 + }, + { + "epoch": 2.695060632948832, + "grad_norm": 1.6321580410003662, + "learning_rate": 2.1525e-06, + "loss": 0.0739, + "step": 2278 + }, + { + "epoch": 2.696243714877255, + "grad_norm": 1.7492852210998535, + "learning_rate": 2.1512500000000004e-06, + "loss": 0.0675, + "step": 2279 + }, + { + "epoch": 2.697426796805679, + "grad_norm": 1.4724619388580322, + "learning_rate": 2.15e-06, + "loss": 0.0693, + "step": 2280 + }, + { + "epoch": 2.698609878734102, + "grad_norm": 1.33844792842865, + "learning_rate": 2.1487500000000004e-06, + "loss": 0.0714, + "step": 2281 + }, + { + "epoch": 2.699792960662526, + "grad_norm": 1.3352513313293457, + "learning_rate": 2.1475e-06, + "loss": 0.0614, + "step": 2282 + }, + { + "epoch": 2.7009760425909493, + "grad_norm": 1.638075590133667, + "learning_rate": 2.1462500000000004e-06, + "loss": 0.0719, + "step": 2283 + }, + { + "epoch": 2.702159124519373, + "grad_norm": 1.4793050289154053, + "learning_rate": 2.1450000000000002e-06, + "loss": 0.0656, + "step": 2284 + }, + { + "epoch": 2.7033422064477968, + "grad_norm": 1.753203272819519, + "learning_rate": 2.1437500000000004e-06, + "loss": 0.0748, + "step": 2285 + }, + { + "epoch": 2.70452528837622, + "grad_norm": 1.5586817264556885, + "learning_rate": 2.1425000000000002e-06, + "loss": 0.0796, + "step": 2286 + }, + { + "epoch": 2.7057083703046434, + "grad_norm": 1.6233011484146118, + "learning_rate": 2.1412500000000005e-06, + "loss": 0.071, + "step": 2287 + }, + { + "epoch": 2.706891452233067, + "grad_norm": 1.795044183731079, + "learning_rate": 2.1400000000000003e-06, + "loss": 0.0715, + "step": 2288 + }, + { + "epoch": 2.708074534161491, + "grad_norm": 1.6033660173416138, + "learning_rate": 2.13875e-06, + "loss": 0.0714, + "step": 2289 + }, + { + "epoch": 2.7092576160899142, + "grad_norm": 2.183138608932495, + "learning_rate": 2.1375000000000003e-06, + "loss": 0.0677, + "step": 2290 + }, + { + "epoch": 2.7104406980183375, + "grad_norm": 1.5675545930862427, + "learning_rate": 2.13625e-06, + "loss": 0.067, + "step": 2291 + }, + { + "epoch": 2.7116237799467613, + "grad_norm": 1.794681429862976, + "learning_rate": 2.1350000000000003e-06, + "loss": 0.0706, + "step": 2292 + }, + { + "epoch": 2.712806861875185, + "grad_norm": 1.730126976966858, + "learning_rate": 2.13375e-06, + "loss": 0.0701, + "step": 2293 + }, + { + "epoch": 2.7139899438036084, + "grad_norm": 1.2990751266479492, + "learning_rate": 2.1325000000000003e-06, + "loss": 0.0707, + "step": 2294 + }, + { + "epoch": 2.7151730257320317, + "grad_norm": 1.4179269075393677, + "learning_rate": 2.13125e-06, + "loss": 0.072, + "step": 2295 + }, + { + "epoch": 2.7163561076604554, + "grad_norm": 1.4814989566802979, + "learning_rate": 2.13e-06, + "loss": 0.0657, + "step": 2296 + }, + { + "epoch": 2.717539189588879, + "grad_norm": 1.54590904712677, + "learning_rate": 2.12875e-06, + "loss": 0.071, + "step": 2297 + }, + { + "epoch": 2.7187222715173025, + "grad_norm": 1.7370691299438477, + "learning_rate": 2.1275e-06, + "loss": 0.0706, + "step": 2298 + }, + { + "epoch": 2.7199053534457263, + "grad_norm": 1.3091175556182861, + "learning_rate": 2.12625e-06, + "loss": 0.0646, + "step": 2299 + }, + { + "epoch": 2.7210884353741496, + "grad_norm": 1.6400442123413086, + "learning_rate": 2.125e-06, + "loss": 0.0728, + "step": 2300 + }, + { + "epoch": 2.7222715173025733, + "grad_norm": 1.6742056608200073, + "learning_rate": 2.12375e-06, + "loss": 0.0708, + "step": 2301 + }, + { + "epoch": 2.7234545992309966, + "grad_norm": 1.6465054750442505, + "learning_rate": 2.1225e-06, + "loss": 0.0679, + "step": 2302 + }, + { + "epoch": 2.7246376811594204, + "grad_norm": 1.574346899986267, + "learning_rate": 2.1212500000000002e-06, + "loss": 0.0733, + "step": 2303 + }, + { + "epoch": 2.7258207630878437, + "grad_norm": 1.8331743478775024, + "learning_rate": 2.12e-06, + "loss": 0.0676, + "step": 2304 + }, + { + "epoch": 2.7270038450162675, + "grad_norm": 1.5618512630462646, + "learning_rate": 2.1187500000000003e-06, + "loss": 0.0637, + "step": 2305 + }, + { + "epoch": 2.728186926944691, + "grad_norm": 1.9284738302230835, + "learning_rate": 2.1175e-06, + "loss": 0.0699, + "step": 2306 + }, + { + "epoch": 2.7293700088731145, + "grad_norm": 1.4501124620437622, + "learning_rate": 2.1162500000000003e-06, + "loss": 0.0705, + "step": 2307 + }, + { + "epoch": 2.730553090801538, + "grad_norm": 1.49224853515625, + "learning_rate": 2.115e-06, + "loss": 0.0668, + "step": 2308 + }, + { + "epoch": 2.7317361727299616, + "grad_norm": 1.7393361330032349, + "learning_rate": 2.1137500000000003e-06, + "loss": 0.0692, + "step": 2309 + }, + { + "epoch": 2.732919254658385, + "grad_norm": 2.2557027339935303, + "learning_rate": 2.1125e-06, + "loss": 0.0801, + "step": 2310 + }, + { + "epoch": 2.7341023365868087, + "grad_norm": 1.4057419300079346, + "learning_rate": 2.1112500000000003e-06, + "loss": 0.067, + "step": 2311 + }, + { + "epoch": 2.7352854185152324, + "grad_norm": 1.528570532798767, + "learning_rate": 2.11e-06, + "loss": 0.0725, + "step": 2312 + }, + { + "epoch": 2.7364685004436557, + "grad_norm": 1.7765111923217773, + "learning_rate": 2.1087500000000004e-06, + "loss": 0.0741, + "step": 2313 + }, + { + "epoch": 2.737651582372079, + "grad_norm": 1.6597504615783691, + "learning_rate": 2.1075e-06, + "loss": 0.0681, + "step": 2314 + }, + { + "epoch": 2.738834664300503, + "grad_norm": 1.766278624534607, + "learning_rate": 2.1062500000000004e-06, + "loss": 0.0811, + "step": 2315 + }, + { + "epoch": 2.7400177462289266, + "grad_norm": 1.8278440237045288, + "learning_rate": 2.105e-06, + "loss": 0.0723, + "step": 2316 + }, + { + "epoch": 2.74120082815735, + "grad_norm": 1.9047292470932007, + "learning_rate": 2.1037500000000004e-06, + "loss": 0.0728, + "step": 2317 + }, + { + "epoch": 2.742383910085773, + "grad_norm": 1.5668994188308716, + "learning_rate": 2.1025e-06, + "loss": 0.0691, + "step": 2318 + }, + { + "epoch": 2.743566992014197, + "grad_norm": 1.4317786693572998, + "learning_rate": 2.1012500000000004e-06, + "loss": 0.0644, + "step": 2319 + }, + { + "epoch": 2.7447500739426207, + "grad_norm": 1.686684489250183, + "learning_rate": 2.1000000000000002e-06, + "loss": 0.0662, + "step": 2320 + }, + { + "epoch": 2.745933155871044, + "grad_norm": 1.5528439283370972, + "learning_rate": 2.09875e-06, + "loss": 0.0728, + "step": 2321 + }, + { + "epoch": 2.7471162377994673, + "grad_norm": 1.880333662033081, + "learning_rate": 2.0975000000000002e-06, + "loss": 0.0728, + "step": 2322 + }, + { + "epoch": 2.748299319727891, + "grad_norm": 1.5328550338745117, + "learning_rate": 2.09625e-06, + "loss": 0.0733, + "step": 2323 + }, + { + "epoch": 2.749482401656315, + "grad_norm": 1.770662784576416, + "learning_rate": 2.0950000000000003e-06, + "loss": 0.074, + "step": 2324 + }, + { + "epoch": 2.750665483584738, + "grad_norm": 1.5119760036468506, + "learning_rate": 2.09375e-06, + "loss": 0.0698, + "step": 2325 + }, + { + "epoch": 2.751848565513162, + "grad_norm": 1.8345869779586792, + "learning_rate": 2.0925000000000003e-06, + "loss": 0.068, + "step": 2326 + }, + { + "epoch": 2.7530316474415852, + "grad_norm": 1.6058146953582764, + "learning_rate": 2.09125e-06, + "loss": 0.0706, + "step": 2327 + }, + { + "epoch": 2.754214729370009, + "grad_norm": 1.740319848060608, + "learning_rate": 2.09e-06, + "loss": 0.065, + "step": 2328 + }, + { + "epoch": 2.7553978112984323, + "grad_norm": 1.5986335277557373, + "learning_rate": 2.08875e-06, + "loss": 0.071, + "step": 2329 + }, + { + "epoch": 2.756580893226856, + "grad_norm": 1.6659438610076904, + "learning_rate": 2.0875e-06, + "loss": 0.0751, + "step": 2330 + }, + { + "epoch": 2.7577639751552794, + "grad_norm": 1.545061707496643, + "learning_rate": 2.08625e-06, + "loss": 0.0681, + "step": 2331 + }, + { + "epoch": 2.758947057083703, + "grad_norm": 1.4119385480880737, + "learning_rate": 2.085e-06, + "loss": 0.0656, + "step": 2332 + }, + { + "epoch": 2.7601301390121264, + "grad_norm": 1.568363904953003, + "learning_rate": 2.08375e-06, + "loss": 0.0752, + "step": 2333 + }, + { + "epoch": 2.76131322094055, + "grad_norm": 1.7086321115493774, + "learning_rate": 2.0825e-06, + "loss": 0.0694, + "step": 2334 + }, + { + "epoch": 2.7624963028689735, + "grad_norm": 1.3216179609298706, + "learning_rate": 2.08125e-06, + "loss": 0.0654, + "step": 2335 + }, + { + "epoch": 2.7636793847973973, + "grad_norm": 1.6458162069320679, + "learning_rate": 2.08e-06, + "loss": 0.0669, + "step": 2336 + }, + { + "epoch": 2.7648624667258206, + "grad_norm": 1.587192416191101, + "learning_rate": 2.0787500000000002e-06, + "loss": 0.0689, + "step": 2337 + }, + { + "epoch": 2.7660455486542443, + "grad_norm": 1.442508339881897, + "learning_rate": 2.0775e-06, + "loss": 0.0717, + "step": 2338 + }, + { + "epoch": 2.767228630582668, + "grad_norm": 1.3573492765426636, + "learning_rate": 2.0762500000000002e-06, + "loss": 0.0691, + "step": 2339 + }, + { + "epoch": 2.7684117125110914, + "grad_norm": 1.405103325843811, + "learning_rate": 2.075e-06, + "loss": 0.0666, + "step": 2340 + }, + { + "epoch": 2.7695947944395147, + "grad_norm": 1.4580464363098145, + "learning_rate": 2.0737500000000003e-06, + "loss": 0.0755, + "step": 2341 + }, + { + "epoch": 2.7707778763679385, + "grad_norm": 1.6686561107635498, + "learning_rate": 2.0725e-06, + "loss": 0.0708, + "step": 2342 + }, + { + "epoch": 2.7719609582963622, + "grad_norm": 1.8269727230072021, + "learning_rate": 2.0712500000000003e-06, + "loss": 0.0719, + "step": 2343 + }, + { + "epoch": 2.7731440402247856, + "grad_norm": 1.5728651285171509, + "learning_rate": 2.07e-06, + "loss": 0.0735, + "step": 2344 + }, + { + "epoch": 2.774327122153209, + "grad_norm": 1.5771796703338623, + "learning_rate": 2.0687500000000003e-06, + "loss": 0.0659, + "step": 2345 + }, + { + "epoch": 2.7755102040816326, + "grad_norm": 1.563912272453308, + "learning_rate": 2.0675e-06, + "loss": 0.0643, + "step": 2346 + }, + { + "epoch": 2.7766932860100564, + "grad_norm": 2.108916997909546, + "learning_rate": 2.0662500000000003e-06, + "loss": 0.0759, + "step": 2347 + }, + { + "epoch": 2.7778763679384797, + "grad_norm": 1.9472790956497192, + "learning_rate": 2.065e-06, + "loss": 0.0697, + "step": 2348 + }, + { + "epoch": 2.779059449866903, + "grad_norm": 1.483688473701477, + "learning_rate": 2.0637500000000004e-06, + "loss": 0.0725, + "step": 2349 + }, + { + "epoch": 2.7802425317953268, + "grad_norm": 1.924587607383728, + "learning_rate": 2.0625e-06, + "loss": 0.0679, + "step": 2350 + }, + { + "epoch": 2.7814256137237505, + "grad_norm": 1.7100998163223267, + "learning_rate": 2.0612500000000004e-06, + "loss": 0.0662, + "step": 2351 + }, + { + "epoch": 2.782608695652174, + "grad_norm": 1.5521677732467651, + "learning_rate": 2.06e-06, + "loss": 0.0685, + "step": 2352 + }, + { + "epoch": 2.7837917775805976, + "grad_norm": 1.555383324623108, + "learning_rate": 2.05875e-06, + "loss": 0.0773, + "step": 2353 + }, + { + "epoch": 2.784974859509021, + "grad_norm": 1.50209379196167, + "learning_rate": 2.0575e-06, + "loss": 0.0631, + "step": 2354 + }, + { + "epoch": 2.7861579414374447, + "grad_norm": 1.7865391969680786, + "learning_rate": 2.05625e-06, + "loss": 0.0732, + "step": 2355 + }, + { + "epoch": 2.787341023365868, + "grad_norm": 1.5945123434066772, + "learning_rate": 2.0550000000000002e-06, + "loss": 0.0651, + "step": 2356 + }, + { + "epoch": 2.7885241052942917, + "grad_norm": 1.7500702142715454, + "learning_rate": 2.05375e-06, + "loss": 0.0678, + "step": 2357 + }, + { + "epoch": 2.789707187222715, + "grad_norm": 1.6575450897216797, + "learning_rate": 2.0525000000000003e-06, + "loss": 0.0769, + "step": 2358 + }, + { + "epoch": 2.790890269151139, + "grad_norm": 1.658270001411438, + "learning_rate": 2.05125e-06, + "loss": 0.0647, + "step": 2359 + }, + { + "epoch": 2.792073351079562, + "grad_norm": 1.5393556356430054, + "learning_rate": 2.05e-06, + "loss": 0.0658, + "step": 2360 + }, + { + "epoch": 2.793256433007986, + "grad_norm": 1.7464255094528198, + "learning_rate": 2.04875e-06, + "loss": 0.0701, + "step": 2361 + }, + { + "epoch": 2.794439514936409, + "grad_norm": 1.427376627922058, + "learning_rate": 2.0475e-06, + "loss": 0.0641, + "step": 2362 + }, + { + "epoch": 2.795622596864833, + "grad_norm": 1.6500639915466309, + "learning_rate": 2.04625e-06, + "loss": 0.0726, + "step": 2363 + }, + { + "epoch": 2.7968056787932563, + "grad_norm": 1.5256954431533813, + "learning_rate": 2.045e-06, + "loss": 0.0705, + "step": 2364 + }, + { + "epoch": 2.79798876072168, + "grad_norm": 1.4015778303146362, + "learning_rate": 2.04375e-06, + "loss": 0.073, + "step": 2365 + }, + { + "epoch": 2.7991718426501038, + "grad_norm": 1.6644686460494995, + "learning_rate": 2.0425e-06, + "loss": 0.0641, + "step": 2366 + }, + { + "epoch": 2.800354924578527, + "grad_norm": 1.4752109050750732, + "learning_rate": 2.04125e-06, + "loss": 0.0693, + "step": 2367 + }, + { + "epoch": 2.8015380065069504, + "grad_norm": 1.4042495489120483, + "learning_rate": 2.04e-06, + "loss": 0.0696, + "step": 2368 + }, + { + "epoch": 2.802721088435374, + "grad_norm": 1.6177476644515991, + "learning_rate": 2.03875e-06, + "loss": 0.0762, + "step": 2369 + }, + { + "epoch": 2.803904170363798, + "grad_norm": 1.3314510583877563, + "learning_rate": 2.0375e-06, + "loss": 0.0672, + "step": 2370 + }, + { + "epoch": 2.8050872522922212, + "grad_norm": 1.3436223268508911, + "learning_rate": 2.03625e-06, + "loss": 0.0656, + "step": 2371 + }, + { + "epoch": 2.8062703342206445, + "grad_norm": 1.8525322675704956, + "learning_rate": 2.035e-06, + "loss": 0.0688, + "step": 2372 + }, + { + "epoch": 2.8074534161490683, + "grad_norm": 1.8247231245040894, + "learning_rate": 2.0337500000000002e-06, + "loss": 0.0712, + "step": 2373 + }, + { + "epoch": 2.808636498077492, + "grad_norm": 1.4588197469711304, + "learning_rate": 2.0325e-06, + "loss": 0.0647, + "step": 2374 + }, + { + "epoch": 2.8098195800059154, + "grad_norm": 1.4315632581710815, + "learning_rate": 2.0312500000000002e-06, + "loss": 0.0649, + "step": 2375 + }, + { + "epoch": 2.8110026619343387, + "grad_norm": 2.0007364749908447, + "learning_rate": 2.0300000000000005e-06, + "loss": 0.0742, + "step": 2376 + }, + { + "epoch": 2.8121857438627624, + "grad_norm": 1.61897611618042, + "learning_rate": 2.0287500000000003e-06, + "loss": 0.0699, + "step": 2377 + }, + { + "epoch": 2.813368825791186, + "grad_norm": 1.4759095907211304, + "learning_rate": 2.0275000000000005e-06, + "loss": 0.0648, + "step": 2378 + }, + { + "epoch": 2.8145519077196095, + "grad_norm": 1.254772663116455, + "learning_rate": 2.0262500000000003e-06, + "loss": 0.0635, + "step": 2379 + }, + { + "epoch": 2.8157349896480333, + "grad_norm": 1.6745940446853638, + "learning_rate": 2.025e-06, + "loss": 0.0666, + "step": 2380 + }, + { + "epoch": 2.8169180715764566, + "grad_norm": 1.575006127357483, + "learning_rate": 2.0237500000000003e-06, + "loss": 0.0668, + "step": 2381 + }, + { + "epoch": 2.8181011535048803, + "grad_norm": 1.8126695156097412, + "learning_rate": 2.0225e-06, + "loss": 0.0807, + "step": 2382 + }, + { + "epoch": 2.8192842354333036, + "grad_norm": 1.5516353845596313, + "learning_rate": 2.0212500000000003e-06, + "loss": 0.0675, + "step": 2383 + }, + { + "epoch": 2.8204673173617274, + "grad_norm": 1.6141889095306396, + "learning_rate": 2.02e-06, + "loss": 0.0676, + "step": 2384 + }, + { + "epoch": 2.8216503992901507, + "grad_norm": 1.626502275466919, + "learning_rate": 2.0187500000000004e-06, + "loss": 0.0732, + "step": 2385 + }, + { + "epoch": 2.8228334812185745, + "grad_norm": 1.7113765478134155, + "learning_rate": 2.0175e-06, + "loss": 0.0696, + "step": 2386 + }, + { + "epoch": 2.824016563146998, + "grad_norm": 1.7847610712051392, + "learning_rate": 2.01625e-06, + "loss": 0.0749, + "step": 2387 + }, + { + "epoch": 2.8251996450754215, + "grad_norm": 1.7121944427490234, + "learning_rate": 2.015e-06, + "loss": 0.0705, + "step": 2388 + }, + { + "epoch": 2.826382727003845, + "grad_norm": 1.5579373836517334, + "learning_rate": 2.01375e-06, + "loss": 0.0673, + "step": 2389 + }, + { + "epoch": 2.8275658089322686, + "grad_norm": 1.552991509437561, + "learning_rate": 2.0125000000000002e-06, + "loss": 0.0642, + "step": 2390 + }, + { + "epoch": 2.828748890860692, + "grad_norm": 1.6552772521972656, + "learning_rate": 2.01125e-06, + "loss": 0.0717, + "step": 2391 + }, + { + "epoch": 2.8299319727891157, + "grad_norm": 1.7462352514266968, + "learning_rate": 2.0100000000000002e-06, + "loss": 0.0682, + "step": 2392 + }, + { + "epoch": 2.8311150547175394, + "grad_norm": 1.5442993640899658, + "learning_rate": 2.00875e-06, + "loss": 0.0746, + "step": 2393 + }, + { + "epoch": 2.8322981366459627, + "grad_norm": 1.7667704820632935, + "learning_rate": 2.0075000000000003e-06, + "loss": 0.0682, + "step": 2394 + }, + { + "epoch": 2.833481218574386, + "grad_norm": 1.4103814363479614, + "learning_rate": 2.00625e-06, + "loss": 0.0727, + "step": 2395 + }, + { + "epoch": 2.83466430050281, + "grad_norm": 1.6469242572784424, + "learning_rate": 2.0050000000000003e-06, + "loss": 0.0706, + "step": 2396 + }, + { + "epoch": 2.8358473824312336, + "grad_norm": 1.5230721235275269, + "learning_rate": 2.00375e-06, + "loss": 0.0709, + "step": 2397 + }, + { + "epoch": 2.837030464359657, + "grad_norm": 1.5549746751785278, + "learning_rate": 2.0025000000000003e-06, + "loss": 0.0717, + "step": 2398 + }, + { + "epoch": 2.83821354628808, + "grad_norm": 1.5702377557754517, + "learning_rate": 2.00125e-06, + "loss": 0.0572, + "step": 2399 + }, + { + "epoch": 2.839396628216504, + "grad_norm": 1.6052407026290894, + "learning_rate": 2.0000000000000003e-06, + "loss": 0.0723, + "step": 2400 + }, + { + "epoch": 2.8405797101449277, + "grad_norm": 1.896070122718811, + "learning_rate": 1.99875e-06, + "loss": 0.0709, + "step": 2401 + }, + { + "epoch": 2.841762792073351, + "grad_norm": 1.688833475112915, + "learning_rate": 1.9975000000000004e-06, + "loss": 0.0681, + "step": 2402 + }, + { + "epoch": 2.842945874001775, + "grad_norm": 1.7648653984069824, + "learning_rate": 1.99625e-06, + "loss": 0.0646, + "step": 2403 + }, + { + "epoch": 2.844128955930198, + "grad_norm": 1.6799945831298828, + "learning_rate": 1.9950000000000004e-06, + "loss": 0.0661, + "step": 2404 + }, + { + "epoch": 2.845312037858622, + "grad_norm": 1.5511871576309204, + "learning_rate": 1.99375e-06, + "loss": 0.0711, + "step": 2405 + }, + { + "epoch": 2.846495119787045, + "grad_norm": 1.8639742136001587, + "learning_rate": 1.9925000000000004e-06, + "loss": 0.0702, + "step": 2406 + }, + { + "epoch": 2.847678201715469, + "grad_norm": 1.360324740409851, + "learning_rate": 1.99125e-06, + "loss": 0.0645, + "step": 2407 + }, + { + "epoch": 2.8488612836438922, + "grad_norm": 1.6489579677581787, + "learning_rate": 1.9900000000000004e-06, + "loss": 0.068, + "step": 2408 + }, + { + "epoch": 2.850044365572316, + "grad_norm": 1.6827703714370728, + "learning_rate": 1.9887500000000002e-06, + "loss": 0.0703, + "step": 2409 + }, + { + "epoch": 2.8512274475007393, + "grad_norm": 1.5007816553115845, + "learning_rate": 1.9875000000000005e-06, + "loss": 0.0669, + "step": 2410 + }, + { + "epoch": 2.852410529429163, + "grad_norm": 1.5588223934173584, + "learning_rate": 1.9862500000000003e-06, + "loss": 0.0619, + "step": 2411 + }, + { + "epoch": 2.8535936113575864, + "grad_norm": 1.927217721939087, + "learning_rate": 1.985e-06, + "loss": 0.0732, + "step": 2412 + }, + { + "epoch": 2.85477669328601, + "grad_norm": 1.6326324939727783, + "learning_rate": 1.9837500000000003e-06, + "loss": 0.0725, + "step": 2413 + }, + { + "epoch": 2.8559597752144334, + "grad_norm": 1.5975621938705444, + "learning_rate": 1.9825e-06, + "loss": 0.0698, + "step": 2414 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 1.4733188152313232, + "learning_rate": 1.9812500000000003e-06, + "loss": 0.0722, + "step": 2415 + }, + { + "epoch": 2.858325939071281, + "grad_norm": 1.7527343034744263, + "learning_rate": 1.98e-06, + "loss": 0.0711, + "step": 2416 + }, + { + "epoch": 2.8595090209997043, + "grad_norm": 1.7432364225387573, + "learning_rate": 1.9787500000000003e-06, + "loss": 0.0644, + "step": 2417 + }, + { + "epoch": 2.8606921029281276, + "grad_norm": 1.6101754903793335, + "learning_rate": 1.9775e-06, + "loss": 0.064, + "step": 2418 + }, + { + "epoch": 2.8618751848565513, + "grad_norm": 1.5554190874099731, + "learning_rate": 1.97625e-06, + "loss": 0.0703, + "step": 2419 + }, + { + "epoch": 2.863058266784975, + "grad_norm": 1.6532433032989502, + "learning_rate": 1.975e-06, + "loss": 0.0719, + "step": 2420 + }, + { + "epoch": 2.8642413487133984, + "grad_norm": 1.4529680013656616, + "learning_rate": 1.97375e-06, + "loss": 0.0645, + "step": 2421 + }, + { + "epoch": 2.8654244306418217, + "grad_norm": 1.3631700277328491, + "learning_rate": 1.9725e-06, + "loss": 0.0634, + "step": 2422 + }, + { + "epoch": 2.8666075125702455, + "grad_norm": 1.7612366676330566, + "learning_rate": 1.97125e-06, + "loss": 0.0651, + "step": 2423 + }, + { + "epoch": 2.8677905944986692, + "grad_norm": 1.6864662170410156, + "learning_rate": 1.97e-06, + "loss": 0.072, + "step": 2424 + }, + { + "epoch": 2.8689736764270926, + "grad_norm": 1.8349568843841553, + "learning_rate": 1.96875e-06, + "loss": 0.0717, + "step": 2425 + }, + { + "epoch": 2.870156758355516, + "grad_norm": 1.5370631217956543, + "learning_rate": 1.9675000000000002e-06, + "loss": 0.0682, + "step": 2426 + }, + { + "epoch": 2.8713398402839396, + "grad_norm": 1.5290476083755493, + "learning_rate": 1.96625e-06, + "loss": 0.0692, + "step": 2427 + }, + { + "epoch": 2.8725229222123634, + "grad_norm": 1.460601568222046, + "learning_rate": 1.9650000000000002e-06, + "loss": 0.0611, + "step": 2428 + }, + { + "epoch": 2.8737060041407867, + "grad_norm": 1.7049473524093628, + "learning_rate": 1.96375e-06, + "loss": 0.07, + "step": 2429 + }, + { + "epoch": 2.8748890860692105, + "grad_norm": 1.5584312677383423, + "learning_rate": 1.9625000000000003e-06, + "loss": 0.0684, + "step": 2430 + }, + { + "epoch": 2.8760721679976338, + "grad_norm": 1.7177371978759766, + "learning_rate": 1.96125e-06, + "loss": 0.0646, + "step": 2431 + }, + { + "epoch": 2.8772552499260575, + "grad_norm": 1.8652653694152832, + "learning_rate": 1.9600000000000003e-06, + "loss": 0.0682, + "step": 2432 + }, + { + "epoch": 2.878438331854481, + "grad_norm": 1.4826210737228394, + "learning_rate": 1.95875e-06, + "loss": 0.0645, + "step": 2433 + }, + { + "epoch": 2.8796214137829046, + "grad_norm": 1.9012829065322876, + "learning_rate": 1.9575000000000003e-06, + "loss": 0.0654, + "step": 2434 + }, + { + "epoch": 2.880804495711328, + "grad_norm": 1.6006711721420288, + "learning_rate": 1.95625e-06, + "loss": 0.0592, + "step": 2435 + }, + { + "epoch": 2.8819875776397517, + "grad_norm": 1.7712098360061646, + "learning_rate": 1.9550000000000003e-06, + "loss": 0.0759, + "step": 2436 + }, + { + "epoch": 2.883170659568175, + "grad_norm": 1.743304967880249, + "learning_rate": 1.95375e-06, + "loss": 0.0738, + "step": 2437 + }, + { + "epoch": 2.8843537414965987, + "grad_norm": 1.6532620191574097, + "learning_rate": 1.9525000000000004e-06, + "loss": 0.0699, + "step": 2438 + }, + { + "epoch": 2.885536823425022, + "grad_norm": 1.3207775354385376, + "learning_rate": 1.95125e-06, + "loss": 0.0663, + "step": 2439 + }, + { + "epoch": 2.886719905353446, + "grad_norm": 1.431752324104309, + "learning_rate": 1.9500000000000004e-06, + "loss": 0.0695, + "step": 2440 + }, + { + "epoch": 2.887902987281869, + "grad_norm": 1.5717995166778564, + "learning_rate": 1.94875e-06, + "loss": 0.0691, + "step": 2441 + }, + { + "epoch": 2.889086069210293, + "grad_norm": 1.7575595378875732, + "learning_rate": 1.9475000000000004e-06, + "loss": 0.0657, + "step": 2442 + }, + { + "epoch": 2.8902691511387166, + "grad_norm": 1.645248532295227, + "learning_rate": 1.9462500000000002e-06, + "loss": 0.0652, + "step": 2443 + }, + { + "epoch": 2.89145223306714, + "grad_norm": 2.034071445465088, + "learning_rate": 1.945e-06, + "loss": 0.0745, + "step": 2444 + }, + { + "epoch": 2.8926353149955633, + "grad_norm": 1.5665355920791626, + "learning_rate": 1.9437500000000002e-06, + "loss": 0.0655, + "step": 2445 + }, + { + "epoch": 2.893818396923987, + "grad_norm": 1.5432153940200806, + "learning_rate": 1.9425e-06, + "loss": 0.0593, + "step": 2446 + }, + { + "epoch": 2.8950014788524108, + "grad_norm": 1.5774388313293457, + "learning_rate": 1.9412500000000003e-06, + "loss": 0.0738, + "step": 2447 + }, + { + "epoch": 2.896184560780834, + "grad_norm": 1.544265866279602, + "learning_rate": 1.94e-06, + "loss": 0.0637, + "step": 2448 + }, + { + "epoch": 2.8973676427092574, + "grad_norm": 1.7072991132736206, + "learning_rate": 1.9387500000000003e-06, + "loss": 0.0797, + "step": 2449 + }, + { + "epoch": 2.898550724637681, + "grad_norm": 1.8152703046798706, + "learning_rate": 1.9375e-06, + "loss": 0.0681, + "step": 2450 + }, + { + "epoch": 2.899733806566105, + "grad_norm": 1.5778100490570068, + "learning_rate": 1.93625e-06, + "loss": 0.0637, + "step": 2451 + }, + { + "epoch": 2.9009168884945282, + "grad_norm": 1.50554621219635, + "learning_rate": 1.935e-06, + "loss": 0.0689, + "step": 2452 + }, + { + "epoch": 2.9020999704229515, + "grad_norm": 1.7500492334365845, + "learning_rate": 1.93375e-06, + "loss": 0.0619, + "step": 2453 + }, + { + "epoch": 2.9032830523513753, + "grad_norm": 1.7047792673110962, + "learning_rate": 1.9325e-06, + "loss": 0.0615, + "step": 2454 + }, + { + "epoch": 2.904466134279799, + "grad_norm": 1.4958536624908447, + "learning_rate": 1.93125e-06, + "loss": 0.069, + "step": 2455 + }, + { + "epoch": 2.9056492162082224, + "grad_norm": 1.4216097593307495, + "learning_rate": 1.93e-06, + "loss": 0.067, + "step": 2456 + }, + { + "epoch": 2.906832298136646, + "grad_norm": 1.3515712022781372, + "learning_rate": 1.92875e-06, + "loss": 0.0665, + "step": 2457 + }, + { + "epoch": 2.9080153800650694, + "grad_norm": 1.7064274549484253, + "learning_rate": 1.9275e-06, + "loss": 0.0678, + "step": 2458 + }, + { + "epoch": 2.909198461993493, + "grad_norm": 1.4822463989257812, + "learning_rate": 1.92625e-06, + "loss": 0.0676, + "step": 2459 + }, + { + "epoch": 2.9103815439219165, + "grad_norm": 1.626153588294983, + "learning_rate": 1.925e-06, + "loss": 0.0682, + "step": 2460 + }, + { + "epoch": 2.9115646258503403, + "grad_norm": 1.7121652364730835, + "learning_rate": 1.92375e-06, + "loss": 0.0675, + "step": 2461 + }, + { + "epoch": 2.9127477077787636, + "grad_norm": 1.6162493228912354, + "learning_rate": 1.9225000000000002e-06, + "loss": 0.0685, + "step": 2462 + }, + { + "epoch": 2.9139307897071873, + "grad_norm": 1.5459274053573608, + "learning_rate": 1.92125e-06, + "loss": 0.0646, + "step": 2463 + }, + { + "epoch": 2.9151138716356106, + "grad_norm": 1.5670764446258545, + "learning_rate": 1.9200000000000003e-06, + "loss": 0.0746, + "step": 2464 + }, + { + "epoch": 2.9162969535640344, + "grad_norm": 1.5584765672683716, + "learning_rate": 1.91875e-06, + "loss": 0.0687, + "step": 2465 + }, + { + "epoch": 2.9174800354924577, + "grad_norm": 1.8651340007781982, + "learning_rate": 1.9175000000000003e-06, + "loss": 0.0717, + "step": 2466 + }, + { + "epoch": 2.9186631174208815, + "grad_norm": 1.623411774635315, + "learning_rate": 1.91625e-06, + "loss": 0.0649, + "step": 2467 + }, + { + "epoch": 2.919846199349305, + "grad_norm": 1.6388531923294067, + "learning_rate": 1.9150000000000003e-06, + "loss": 0.0719, + "step": 2468 + }, + { + "epoch": 2.9210292812777285, + "grad_norm": 1.4986459016799927, + "learning_rate": 1.91375e-06, + "loss": 0.0653, + "step": 2469 + }, + { + "epoch": 2.9222123632061523, + "grad_norm": 1.4050724506378174, + "learning_rate": 1.9125000000000003e-06, + "loss": 0.072, + "step": 2470 + }, + { + "epoch": 2.9233954451345756, + "grad_norm": 1.807381272315979, + "learning_rate": 1.91125e-06, + "loss": 0.0708, + "step": 2471 + }, + { + "epoch": 2.924578527062999, + "grad_norm": 1.8097127676010132, + "learning_rate": 1.9100000000000003e-06, + "loss": 0.0688, + "step": 2472 + }, + { + "epoch": 2.9257616089914227, + "grad_norm": 1.526466965675354, + "learning_rate": 1.90875e-06, + "loss": 0.0667, + "step": 2473 + }, + { + "epoch": 2.9269446909198464, + "grad_norm": 1.5119761228561401, + "learning_rate": 1.9075000000000004e-06, + "loss": 0.0649, + "step": 2474 + }, + { + "epoch": 2.9281277728482697, + "grad_norm": 1.5706753730773926, + "learning_rate": 1.90625e-06, + "loss": 0.071, + "step": 2475 + }, + { + "epoch": 2.929310854776693, + "grad_norm": 1.4053676128387451, + "learning_rate": 1.9050000000000002e-06, + "loss": 0.0644, + "step": 2476 + }, + { + "epoch": 2.930493936705117, + "grad_norm": 1.5823434591293335, + "learning_rate": 1.90375e-06, + "loss": 0.0638, + "step": 2477 + }, + { + "epoch": 2.9316770186335406, + "grad_norm": 1.453567385673523, + "learning_rate": 1.9025000000000002e-06, + "loss": 0.0635, + "step": 2478 + }, + { + "epoch": 2.932860100561964, + "grad_norm": 1.5962657928466797, + "learning_rate": 1.90125e-06, + "loss": 0.0675, + "step": 2479 + }, + { + "epoch": 2.934043182490387, + "grad_norm": 1.418526291847229, + "learning_rate": 1.9000000000000002e-06, + "loss": 0.065, + "step": 2480 + }, + { + "epoch": 2.935226264418811, + "grad_norm": 1.5755441188812256, + "learning_rate": 1.89875e-06, + "loss": 0.0664, + "step": 2481 + }, + { + "epoch": 2.9364093463472347, + "grad_norm": 2.1122994422912598, + "learning_rate": 1.8975000000000003e-06, + "loss": 0.074, + "step": 2482 + }, + { + "epoch": 2.937592428275658, + "grad_norm": 1.7100212574005127, + "learning_rate": 1.89625e-06, + "loss": 0.0693, + "step": 2483 + }, + { + "epoch": 2.938775510204082, + "grad_norm": 1.6845777034759521, + "learning_rate": 1.895e-06, + "loss": 0.0677, + "step": 2484 + }, + { + "epoch": 2.939958592132505, + "grad_norm": 1.4111323356628418, + "learning_rate": 1.89375e-06, + "loss": 0.0659, + "step": 2485 + }, + { + "epoch": 2.941141674060929, + "grad_norm": 1.3746331930160522, + "learning_rate": 1.8925e-06, + "loss": 0.0641, + "step": 2486 + }, + { + "epoch": 2.942324755989352, + "grad_norm": 1.542730450630188, + "learning_rate": 1.89125e-06, + "loss": 0.0652, + "step": 2487 + }, + { + "epoch": 2.943507837917776, + "grad_norm": 1.6250323057174683, + "learning_rate": 1.8900000000000001e-06, + "loss": 0.068, + "step": 2488 + }, + { + "epoch": 2.9446909198461992, + "grad_norm": 1.651515007019043, + "learning_rate": 1.8887500000000001e-06, + "loss": 0.0829, + "step": 2489 + }, + { + "epoch": 2.945874001774623, + "grad_norm": 1.6565111875534058, + "learning_rate": 1.8875000000000001e-06, + "loss": 0.0678, + "step": 2490 + }, + { + "epoch": 2.9470570837030463, + "grad_norm": 1.4849125146865845, + "learning_rate": 1.88625e-06, + "loss": 0.0651, + "step": 2491 + }, + { + "epoch": 2.94824016563147, + "grad_norm": 1.3929630517959595, + "learning_rate": 1.8850000000000002e-06, + "loss": 0.0708, + "step": 2492 + }, + { + "epoch": 2.9494232475598934, + "grad_norm": 1.4340704679489136, + "learning_rate": 1.88375e-06, + "loss": 0.0655, + "step": 2493 + }, + { + "epoch": 2.950606329488317, + "grad_norm": 1.6422781944274902, + "learning_rate": 1.8825000000000002e-06, + "loss": 0.073, + "step": 2494 + }, + { + "epoch": 2.9517894114167404, + "grad_norm": 1.732357382774353, + "learning_rate": 1.88125e-06, + "loss": 0.0705, + "step": 2495 + }, + { + "epoch": 2.952972493345164, + "grad_norm": 1.5493820905685425, + "learning_rate": 1.8800000000000002e-06, + "loss": 0.0644, + "step": 2496 + }, + { + "epoch": 2.954155575273588, + "grad_norm": 1.5499510765075684, + "learning_rate": 1.87875e-06, + "loss": 0.0687, + "step": 2497 + }, + { + "epoch": 2.9553386572020113, + "grad_norm": 1.5106990337371826, + "learning_rate": 1.8775000000000002e-06, + "loss": 0.0664, + "step": 2498 + }, + { + "epoch": 2.9565217391304346, + "grad_norm": 1.796994924545288, + "learning_rate": 1.87625e-06, + "loss": 0.0717, + "step": 2499 + }, + { + "epoch": 2.9577048210588583, + "grad_norm": 1.6729393005371094, + "learning_rate": 1.8750000000000003e-06, + "loss": 0.0703, + "step": 2500 + }, + { + "epoch": 2.958887902987282, + "grad_norm": 1.7344276905059814, + "learning_rate": 1.8737500000000003e-06, + "loss": 0.0728, + "step": 2501 + }, + { + "epoch": 2.9600709849157054, + "grad_norm": 1.6145223379135132, + "learning_rate": 1.8725e-06, + "loss": 0.069, + "step": 2502 + }, + { + "epoch": 2.9612540668441287, + "grad_norm": 1.7349733114242554, + "learning_rate": 1.8712500000000003e-06, + "loss": 0.0685, + "step": 2503 + }, + { + "epoch": 2.9624371487725525, + "grad_norm": 1.5592211484909058, + "learning_rate": 1.87e-06, + "loss": 0.068, + "step": 2504 + }, + { + "epoch": 2.9636202307009762, + "grad_norm": 1.6399497985839844, + "learning_rate": 1.8687500000000003e-06, + "loss": 0.0675, + "step": 2505 + }, + { + "epoch": 2.9648033126293996, + "grad_norm": 1.4416548013687134, + "learning_rate": 1.8675000000000001e-06, + "loss": 0.0601, + "step": 2506 + }, + { + "epoch": 2.965986394557823, + "grad_norm": 1.6833360195159912, + "learning_rate": 1.8662500000000003e-06, + "loss": 0.0747, + "step": 2507 + }, + { + "epoch": 2.9671694764862466, + "grad_norm": 1.4294401407241821, + "learning_rate": 1.8650000000000001e-06, + "loss": 0.0718, + "step": 2508 + }, + { + "epoch": 2.9683525584146704, + "grad_norm": 1.6642833948135376, + "learning_rate": 1.8637500000000004e-06, + "loss": 0.0644, + "step": 2509 + }, + { + "epoch": 2.9695356403430937, + "grad_norm": 1.6667840480804443, + "learning_rate": 1.8625000000000002e-06, + "loss": 0.0733, + "step": 2510 + }, + { + "epoch": 2.9707187222715175, + "grad_norm": 1.5777897834777832, + "learning_rate": 1.8612500000000002e-06, + "loss": 0.0696, + "step": 2511 + }, + { + "epoch": 2.9719018041999408, + "grad_norm": 1.4742311239242554, + "learning_rate": 1.8600000000000002e-06, + "loss": 0.0744, + "step": 2512 + }, + { + "epoch": 2.9730848861283645, + "grad_norm": 1.709455966949463, + "learning_rate": 1.8587500000000002e-06, + "loss": 0.0724, + "step": 2513 + }, + { + "epoch": 2.974267968056788, + "grad_norm": 1.4900271892547607, + "learning_rate": 1.8575000000000002e-06, + "loss": 0.0665, + "step": 2514 + }, + { + "epoch": 2.9754510499852116, + "grad_norm": 1.5196034908294678, + "learning_rate": 1.8562500000000002e-06, + "loss": 0.0704, + "step": 2515 + }, + { + "epoch": 2.976634131913635, + "grad_norm": 1.7550339698791504, + "learning_rate": 1.8550000000000002e-06, + "loss": 0.0695, + "step": 2516 + }, + { + "epoch": 2.9778172138420587, + "grad_norm": 1.6801080703735352, + "learning_rate": 1.8537500000000003e-06, + "loss": 0.066, + "step": 2517 + }, + { + "epoch": 2.979000295770482, + "grad_norm": 1.7404299974441528, + "learning_rate": 1.8525e-06, + "loss": 0.0658, + "step": 2518 + }, + { + "epoch": 2.9801833776989057, + "grad_norm": 2.022076368331909, + "learning_rate": 1.8512500000000003e-06, + "loss": 0.0657, + "step": 2519 + }, + { + "epoch": 2.981366459627329, + "grad_norm": 1.74867582321167, + "learning_rate": 1.85e-06, + "loss": 0.0708, + "step": 2520 + }, + { + "epoch": 2.982549541555753, + "grad_norm": 1.9249935150146484, + "learning_rate": 1.8487500000000003e-06, + "loss": 0.065, + "step": 2521 + }, + { + "epoch": 2.983732623484176, + "grad_norm": 1.7662882804870605, + "learning_rate": 1.8475e-06, + "loss": 0.0729, + "step": 2522 + }, + { + "epoch": 2.9849157054126, + "grad_norm": 1.539586067199707, + "learning_rate": 1.8462500000000003e-06, + "loss": 0.0617, + "step": 2523 + }, + { + "epoch": 2.9860987873410236, + "grad_norm": 1.6438349485397339, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.0705, + "step": 2524 + }, + { + "epoch": 2.987281869269447, + "grad_norm": 1.5068515539169312, + "learning_rate": 1.8437500000000003e-06, + "loss": 0.066, + "step": 2525 + }, + { + "epoch": 2.9884649511978703, + "grad_norm": 1.6392298936843872, + "learning_rate": 1.8425000000000001e-06, + "loss": 0.0705, + "step": 2526 + }, + { + "epoch": 2.989648033126294, + "grad_norm": 1.6790026426315308, + "learning_rate": 1.8412500000000002e-06, + "loss": 0.0686, + "step": 2527 + }, + { + "epoch": 2.9908311150547178, + "grad_norm": 1.8331384658813477, + "learning_rate": 1.8400000000000002e-06, + "loss": 0.0679, + "step": 2528 + }, + { + "epoch": 2.992014196983141, + "grad_norm": 1.577783226966858, + "learning_rate": 1.8387500000000002e-06, + "loss": 0.0761, + "step": 2529 + }, + { + "epoch": 2.9931972789115644, + "grad_norm": 1.9697670936584473, + "learning_rate": 1.8375000000000002e-06, + "loss": 0.074, + "step": 2530 + }, + { + "epoch": 2.994380360839988, + "grad_norm": 1.5967445373535156, + "learning_rate": 1.8362500000000002e-06, + "loss": 0.0667, + "step": 2531 + }, + { + "epoch": 2.995563442768412, + "grad_norm": 1.6976232528686523, + "learning_rate": 1.8350000000000002e-06, + "loss": 0.0679, + "step": 2532 + }, + { + "epoch": 2.9967465246968352, + "grad_norm": 1.9708269834518433, + "learning_rate": 1.8337500000000002e-06, + "loss": 0.0628, + "step": 2533 + }, + { + "epoch": 2.9979296066252585, + "grad_norm": 1.6069588661193848, + "learning_rate": 1.8325e-06, + "loss": 0.0662, + "step": 2534 + }, + { + "epoch": 2.9991126885536823, + "grad_norm": 1.5127321481704712, + "learning_rate": 1.8312500000000003e-06, + "loss": 0.076, + "step": 2535 + }, + { + "epoch": 3.000295770482106, + "grad_norm": 1.1114705801010132, + "learning_rate": 1.83e-06, + "loss": 0.0595, + "step": 2536 + }, + { + "epoch": 3.0014788524105294, + "grad_norm": 0.9029242396354675, + "learning_rate": 1.8287500000000003e-06, + "loss": 0.0337, + "step": 2537 + }, + { + "epoch": 3.002661934338953, + "grad_norm": 1.304062843322754, + "learning_rate": 1.8275e-06, + "loss": 0.034, + "step": 2538 + }, + { + "epoch": 3.0038450162673764, + "grad_norm": 1.2144815921783447, + "learning_rate": 1.8262500000000003e-06, + "loss": 0.0376, + "step": 2539 + }, + { + "epoch": 3.0050280981958, + "grad_norm": 1.2010996341705322, + "learning_rate": 1.825e-06, + "loss": 0.0351, + "step": 2540 + }, + { + "epoch": 3.0062111801242235, + "grad_norm": 1.246021032333374, + "learning_rate": 1.8237500000000003e-06, + "loss": 0.0333, + "step": 2541 + }, + { + "epoch": 3.0073942620526473, + "grad_norm": 1.3628509044647217, + "learning_rate": 1.8225000000000001e-06, + "loss": 0.0341, + "step": 2542 + }, + { + "epoch": 3.0085773439810706, + "grad_norm": 2.1943373680114746, + "learning_rate": 1.8212500000000001e-06, + "loss": 0.0353, + "step": 2543 + }, + { + "epoch": 3.0097604259094943, + "grad_norm": 1.4251822233200073, + "learning_rate": 1.8200000000000002e-06, + "loss": 0.0336, + "step": 2544 + }, + { + "epoch": 3.0109435078379176, + "grad_norm": 1.390965223312378, + "learning_rate": 1.8187500000000002e-06, + "loss": 0.0353, + "step": 2545 + }, + { + "epoch": 3.0121265897663414, + "grad_norm": 1.6122360229492188, + "learning_rate": 1.8175000000000002e-06, + "loss": 0.034, + "step": 2546 + }, + { + "epoch": 3.0133096716947647, + "grad_norm": 1.3799641132354736, + "learning_rate": 1.8162500000000002e-06, + "loss": 0.0334, + "step": 2547 + }, + { + "epoch": 3.0144927536231885, + "grad_norm": 1.3610409498214722, + "learning_rate": 1.8150000000000002e-06, + "loss": 0.035, + "step": 2548 + }, + { + "epoch": 3.015675835551612, + "grad_norm": 1.3110113143920898, + "learning_rate": 1.8137500000000002e-06, + "loss": 0.0345, + "step": 2549 + }, + { + "epoch": 3.0168589174800355, + "grad_norm": 1.5193076133728027, + "learning_rate": 1.8125e-06, + "loss": 0.0355, + "step": 2550 + }, + { + "epoch": 3.018041999408459, + "grad_norm": 1.4479187726974487, + "learning_rate": 1.8112500000000002e-06, + "loss": 0.0339, + "step": 2551 + }, + { + "epoch": 3.0192250813368826, + "grad_norm": 1.4371201992034912, + "learning_rate": 1.81e-06, + "loss": 0.0344, + "step": 2552 + }, + { + "epoch": 3.020408163265306, + "grad_norm": 1.4891449213027954, + "learning_rate": 1.8087500000000003e-06, + "loss": 0.0347, + "step": 2553 + }, + { + "epoch": 3.0215912451937297, + "grad_norm": 1.7040371894836426, + "learning_rate": 1.8075e-06, + "loss": 0.0347, + "step": 2554 + }, + { + "epoch": 3.022774327122153, + "grad_norm": 1.3031361103057861, + "learning_rate": 1.8062500000000003e-06, + "loss": 0.0373, + "step": 2555 + }, + { + "epoch": 3.0239574090505768, + "grad_norm": 1.484960675239563, + "learning_rate": 1.805e-06, + "loss": 0.038, + "step": 2556 + }, + { + "epoch": 3.0251404909790005, + "grad_norm": 1.4752142429351807, + "learning_rate": 1.8037500000000003e-06, + "loss": 0.0338, + "step": 2557 + }, + { + "epoch": 3.026323572907424, + "grad_norm": 1.3498085737228394, + "learning_rate": 1.8025000000000001e-06, + "loss": 0.0353, + "step": 2558 + }, + { + "epoch": 3.0275066548358476, + "grad_norm": 1.6069928407669067, + "learning_rate": 1.8012500000000001e-06, + "loss": 0.0367, + "step": 2559 + }, + { + "epoch": 3.028689736764271, + "grad_norm": 1.3902969360351562, + "learning_rate": 1.8000000000000001e-06, + "loss": 0.0378, + "step": 2560 + }, + { + "epoch": 3.0298728186926946, + "grad_norm": 1.5500601530075073, + "learning_rate": 1.7987500000000001e-06, + "loss": 0.0355, + "step": 2561 + }, + { + "epoch": 3.031055900621118, + "grad_norm": 1.239907145500183, + "learning_rate": 1.7975000000000002e-06, + "loss": 0.032, + "step": 2562 + }, + { + "epoch": 3.0322389825495417, + "grad_norm": 1.148078441619873, + "learning_rate": 1.7962500000000002e-06, + "loss": 0.0324, + "step": 2563 + }, + { + "epoch": 3.033422064477965, + "grad_norm": 1.1643887758255005, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.0356, + "step": 2564 + }, + { + "epoch": 3.034605146406389, + "grad_norm": 1.6064841747283936, + "learning_rate": 1.7937500000000002e-06, + "loss": 0.0383, + "step": 2565 + }, + { + "epoch": 3.035788228334812, + "grad_norm": 1.8022316694259644, + "learning_rate": 1.7925e-06, + "loss": 0.036, + "step": 2566 + }, + { + "epoch": 3.036971310263236, + "grad_norm": 1.5448108911514282, + "learning_rate": 1.7912500000000002e-06, + "loss": 0.0313, + "step": 2567 + }, + { + "epoch": 3.038154392191659, + "grad_norm": 1.1204724311828613, + "learning_rate": 1.79e-06, + "loss": 0.0354, + "step": 2568 + }, + { + "epoch": 3.039337474120083, + "grad_norm": 1.5587433576583862, + "learning_rate": 1.7887500000000002e-06, + "loss": 0.0309, + "step": 2569 + }, + { + "epoch": 3.0405205560485062, + "grad_norm": 1.4279117584228516, + "learning_rate": 1.7875e-06, + "loss": 0.0329, + "step": 2570 + }, + { + "epoch": 3.04170363797693, + "grad_norm": 1.7578426599502563, + "learning_rate": 1.7862500000000003e-06, + "loss": 0.0338, + "step": 2571 + }, + { + "epoch": 3.0428867199053533, + "grad_norm": 1.5974630117416382, + "learning_rate": 1.785e-06, + "loss": 0.0348, + "step": 2572 + }, + { + "epoch": 3.044069801833777, + "grad_norm": 1.5781141519546509, + "learning_rate": 1.7837500000000003e-06, + "loss": 0.0369, + "step": 2573 + }, + { + "epoch": 3.0452528837622004, + "grad_norm": 1.5280400514602661, + "learning_rate": 1.7825e-06, + "loss": 0.0352, + "step": 2574 + }, + { + "epoch": 3.046435965690624, + "grad_norm": 2.2156310081481934, + "learning_rate": 1.78125e-06, + "loss": 0.0371, + "step": 2575 + }, + { + "epoch": 3.0476190476190474, + "grad_norm": 1.6461681127548218, + "learning_rate": 1.7800000000000001e-06, + "loss": 0.0357, + "step": 2576 + }, + { + "epoch": 3.048802129547471, + "grad_norm": 1.7484053373336792, + "learning_rate": 1.7787500000000001e-06, + "loss": 0.0337, + "step": 2577 + }, + { + "epoch": 3.0499852114758945, + "grad_norm": 1.3296630382537842, + "learning_rate": 1.7775000000000001e-06, + "loss": 0.0336, + "step": 2578 + }, + { + "epoch": 3.0511682934043183, + "grad_norm": 1.579818844795227, + "learning_rate": 1.7762500000000001e-06, + "loss": 0.0389, + "step": 2579 + }, + { + "epoch": 3.0523513753327416, + "grad_norm": 1.2949336767196655, + "learning_rate": 1.7750000000000002e-06, + "loss": 0.0346, + "step": 2580 + }, + { + "epoch": 3.0535344572611653, + "grad_norm": 1.3661082983016968, + "learning_rate": 1.7737500000000002e-06, + "loss": 0.0315, + "step": 2581 + }, + { + "epoch": 3.054717539189589, + "grad_norm": 1.477048397064209, + "learning_rate": 1.7725e-06, + "loss": 0.0338, + "step": 2582 + }, + { + "epoch": 3.0559006211180124, + "grad_norm": 1.4536473751068115, + "learning_rate": 1.7712500000000002e-06, + "loss": 0.0302, + "step": 2583 + }, + { + "epoch": 3.057083703046436, + "grad_norm": 1.1080058813095093, + "learning_rate": 1.77e-06, + "loss": 0.0304, + "step": 2584 + }, + { + "epoch": 3.0582667849748595, + "grad_norm": 1.4379875659942627, + "learning_rate": 1.7687500000000002e-06, + "loss": 0.032, + "step": 2585 + }, + { + "epoch": 3.0594498669032832, + "grad_norm": 1.7263069152832031, + "learning_rate": 1.7675e-06, + "loss": 0.0361, + "step": 2586 + }, + { + "epoch": 3.0606329488317066, + "grad_norm": 1.5186476707458496, + "learning_rate": 1.7662500000000002e-06, + "loss": 0.032, + "step": 2587 + }, + { + "epoch": 3.0618160307601303, + "grad_norm": 1.5951355695724487, + "learning_rate": 1.765e-06, + "loss": 0.0328, + "step": 2588 + }, + { + "epoch": 3.0629991126885536, + "grad_norm": 1.340288758277893, + "learning_rate": 1.7637500000000003e-06, + "loss": 0.0382, + "step": 2589 + }, + { + "epoch": 3.0641821946169774, + "grad_norm": 1.5069451332092285, + "learning_rate": 1.7625e-06, + "loss": 0.0321, + "step": 2590 + }, + { + "epoch": 3.0653652765454007, + "grad_norm": 1.6718896627426147, + "learning_rate": 1.76125e-06, + "loss": 0.0339, + "step": 2591 + }, + { + "epoch": 3.0665483584738245, + "grad_norm": 1.3780766725540161, + "learning_rate": 1.76e-06, + "loss": 0.0355, + "step": 2592 + }, + { + "epoch": 3.0677314404022478, + "grad_norm": 1.495343804359436, + "learning_rate": 1.75875e-06, + "loss": 0.0349, + "step": 2593 + }, + { + "epoch": 3.0689145223306715, + "grad_norm": 1.5244766473770142, + "learning_rate": 1.7575000000000001e-06, + "loss": 0.0327, + "step": 2594 + }, + { + "epoch": 3.070097604259095, + "grad_norm": 1.6871026754379272, + "learning_rate": 1.7562500000000001e-06, + "loss": 0.0348, + "step": 2595 + }, + { + "epoch": 3.0712806861875186, + "grad_norm": 1.355095386505127, + "learning_rate": 1.7550000000000001e-06, + "loss": 0.0327, + "step": 2596 + }, + { + "epoch": 3.072463768115942, + "grad_norm": 1.577634572982788, + "learning_rate": 1.7537500000000001e-06, + "loss": 0.0357, + "step": 2597 + }, + { + "epoch": 3.0736468500443657, + "grad_norm": 1.5134474039077759, + "learning_rate": 1.7525e-06, + "loss": 0.0306, + "step": 2598 + }, + { + "epoch": 3.074829931972789, + "grad_norm": 1.663620948791504, + "learning_rate": 1.7512500000000002e-06, + "loss": 0.0342, + "step": 2599 + }, + { + "epoch": 3.0760130139012127, + "grad_norm": 1.7172060012817383, + "learning_rate": 1.75e-06, + "loss": 0.036, + "step": 2600 + }, + { + "epoch": 3.077196095829636, + "grad_norm": 1.2387088537216187, + "learning_rate": 1.7487500000000002e-06, + "loss": 0.0356, + "step": 2601 + }, + { + "epoch": 3.07837917775806, + "grad_norm": 1.4813896417617798, + "learning_rate": 1.7475e-06, + "loss": 0.0357, + "step": 2602 + }, + { + "epoch": 3.079562259686483, + "grad_norm": 1.3667938709259033, + "learning_rate": 1.7462500000000002e-06, + "loss": 0.0337, + "step": 2603 + }, + { + "epoch": 3.080745341614907, + "grad_norm": 1.3202913999557495, + "learning_rate": 1.745e-06, + "loss": 0.0333, + "step": 2604 + }, + { + "epoch": 3.08192842354333, + "grad_norm": 1.1989960670471191, + "learning_rate": 1.7437500000000002e-06, + "loss": 0.0331, + "step": 2605 + }, + { + "epoch": 3.083111505471754, + "grad_norm": 1.796905279159546, + "learning_rate": 1.7425e-06, + "loss": 0.0377, + "step": 2606 + }, + { + "epoch": 3.0842945874001773, + "grad_norm": 1.6856011152267456, + "learning_rate": 1.74125e-06, + "loss": 0.0334, + "step": 2607 + }, + { + "epoch": 3.085477669328601, + "grad_norm": 1.2994053363800049, + "learning_rate": 1.74e-06, + "loss": 0.0355, + "step": 2608 + }, + { + "epoch": 3.0866607512570248, + "grad_norm": 1.5401818752288818, + "learning_rate": 1.73875e-06, + "loss": 0.0352, + "step": 2609 + }, + { + "epoch": 3.087843833185448, + "grad_norm": 1.85686194896698, + "learning_rate": 1.7375e-06, + "loss": 0.0328, + "step": 2610 + }, + { + "epoch": 3.089026915113872, + "grad_norm": 1.2153459787368774, + "learning_rate": 1.73625e-06, + "loss": 0.0292, + "step": 2611 + }, + { + "epoch": 3.090209997042295, + "grad_norm": 1.605089545249939, + "learning_rate": 1.7350000000000001e-06, + "loss": 0.0356, + "step": 2612 + }, + { + "epoch": 3.091393078970719, + "grad_norm": 1.241175889968872, + "learning_rate": 1.7337500000000001e-06, + "loss": 0.0315, + "step": 2613 + }, + { + "epoch": 3.0925761608991422, + "grad_norm": 1.4177995920181274, + "learning_rate": 1.7325e-06, + "loss": 0.0426, + "step": 2614 + }, + { + "epoch": 3.093759242827566, + "grad_norm": 1.6079317331314087, + "learning_rate": 1.7312500000000002e-06, + "loss": 0.0353, + "step": 2615 + }, + { + "epoch": 3.0949423247559893, + "grad_norm": 1.4533647298812866, + "learning_rate": 1.73e-06, + "loss": 0.0333, + "step": 2616 + }, + { + "epoch": 3.096125406684413, + "grad_norm": 1.3900617361068726, + "learning_rate": 1.7287500000000002e-06, + "loss": 0.0357, + "step": 2617 + }, + { + "epoch": 3.0973084886128364, + "grad_norm": 1.3681914806365967, + "learning_rate": 1.7275e-06, + "loss": 0.0305, + "step": 2618 + }, + { + "epoch": 3.09849157054126, + "grad_norm": 1.4883403778076172, + "learning_rate": 1.7262500000000002e-06, + "loss": 0.0327, + "step": 2619 + }, + { + "epoch": 3.0996746524696834, + "grad_norm": 1.5963979959487915, + "learning_rate": 1.725e-06, + "loss": 0.0358, + "step": 2620 + }, + { + "epoch": 3.100857734398107, + "grad_norm": 1.5026475191116333, + "learning_rate": 1.7237500000000002e-06, + "loss": 0.0334, + "step": 2621 + }, + { + "epoch": 3.1020408163265305, + "grad_norm": 1.4265025854110718, + "learning_rate": 1.7225e-06, + "loss": 0.0378, + "step": 2622 + }, + { + "epoch": 3.1032238982549543, + "grad_norm": 1.3758147954940796, + "learning_rate": 1.72125e-06, + "loss": 0.0332, + "step": 2623 + }, + { + "epoch": 3.1044069801833776, + "grad_norm": 1.903786301612854, + "learning_rate": 1.72e-06, + "loss": 0.0381, + "step": 2624 + }, + { + "epoch": 3.1055900621118013, + "grad_norm": 1.760132074356079, + "learning_rate": 1.71875e-06, + "loss": 0.0423, + "step": 2625 + }, + { + "epoch": 3.1067731440402246, + "grad_norm": 1.4317656755447388, + "learning_rate": 1.7175000000000003e-06, + "loss": 0.0359, + "step": 2626 + }, + { + "epoch": 3.1079562259686484, + "grad_norm": 1.2694700956344604, + "learning_rate": 1.71625e-06, + "loss": 0.0366, + "step": 2627 + }, + { + "epoch": 3.1091393078970717, + "grad_norm": 1.5842773914337158, + "learning_rate": 1.7150000000000003e-06, + "loss": 0.04, + "step": 2628 + }, + { + "epoch": 3.1103223898254955, + "grad_norm": 1.7515537738800049, + "learning_rate": 1.7137500000000001e-06, + "loss": 0.0366, + "step": 2629 + }, + { + "epoch": 3.111505471753919, + "grad_norm": 1.564864993095398, + "learning_rate": 1.7125000000000003e-06, + "loss": 0.032, + "step": 2630 + }, + { + "epoch": 3.1126885536823425, + "grad_norm": 1.6075620651245117, + "learning_rate": 1.7112500000000001e-06, + "loss": 0.0347, + "step": 2631 + }, + { + "epoch": 3.113871635610766, + "grad_norm": 1.308459758758545, + "learning_rate": 1.7100000000000004e-06, + "loss": 0.0315, + "step": 2632 + }, + { + "epoch": 3.1150547175391896, + "grad_norm": 1.1863787174224854, + "learning_rate": 1.7087500000000002e-06, + "loss": 0.0351, + "step": 2633 + }, + { + "epoch": 3.116237799467613, + "grad_norm": 1.518088936805725, + "learning_rate": 1.7075000000000002e-06, + "loss": 0.0338, + "step": 2634 + }, + { + "epoch": 3.1174208813960367, + "grad_norm": 1.1476041078567505, + "learning_rate": 1.7062500000000002e-06, + "loss": 0.0356, + "step": 2635 + }, + { + "epoch": 3.1186039633244604, + "grad_norm": 1.527772307395935, + "learning_rate": 1.7050000000000002e-06, + "loss": 0.0332, + "step": 2636 + }, + { + "epoch": 3.1197870452528838, + "grad_norm": 1.3114049434661865, + "learning_rate": 1.7037500000000002e-06, + "loss": 0.0404, + "step": 2637 + }, + { + "epoch": 3.1209701271813075, + "grad_norm": 1.3765316009521484, + "learning_rate": 1.7025000000000002e-06, + "loss": 0.0323, + "step": 2638 + }, + { + "epoch": 3.122153209109731, + "grad_norm": 1.6492255926132202, + "learning_rate": 1.7012500000000002e-06, + "loss": 0.0327, + "step": 2639 + }, + { + "epoch": 3.1233362910381546, + "grad_norm": 1.2614279985427856, + "learning_rate": 1.7000000000000002e-06, + "loss": 0.0326, + "step": 2640 + }, + { + "epoch": 3.124519372966578, + "grad_norm": 1.4618563652038574, + "learning_rate": 1.69875e-06, + "loss": 0.0363, + "step": 2641 + }, + { + "epoch": 3.1257024548950016, + "grad_norm": 1.6708273887634277, + "learning_rate": 1.6975000000000003e-06, + "loss": 0.0322, + "step": 2642 + }, + { + "epoch": 3.126885536823425, + "grad_norm": 1.7514479160308838, + "learning_rate": 1.69625e-06, + "loss": 0.0365, + "step": 2643 + }, + { + "epoch": 3.1280686187518487, + "grad_norm": 2.040926694869995, + "learning_rate": 1.6950000000000003e-06, + "loss": 0.036, + "step": 2644 + }, + { + "epoch": 3.129251700680272, + "grad_norm": 1.4218146800994873, + "learning_rate": 1.69375e-06, + "loss": 0.0334, + "step": 2645 + }, + { + "epoch": 3.130434782608696, + "grad_norm": 1.4094136953353882, + "learning_rate": 1.6925000000000003e-06, + "loss": 0.0331, + "step": 2646 + }, + { + "epoch": 3.131617864537119, + "grad_norm": 1.5134646892547607, + "learning_rate": 1.6912500000000001e-06, + "loss": 0.0393, + "step": 2647 + }, + { + "epoch": 3.132800946465543, + "grad_norm": 1.373430609703064, + "learning_rate": 1.6900000000000003e-06, + "loss": 0.0351, + "step": 2648 + }, + { + "epoch": 3.133984028393966, + "grad_norm": 1.5034221410751343, + "learning_rate": 1.6887500000000001e-06, + "loss": 0.035, + "step": 2649 + }, + { + "epoch": 3.13516711032239, + "grad_norm": 2.0055410861968994, + "learning_rate": 1.6875000000000001e-06, + "loss": 0.0352, + "step": 2650 + }, + { + "epoch": 3.1363501922508132, + "grad_norm": 1.4226932525634766, + "learning_rate": 1.6862500000000002e-06, + "loss": 0.0327, + "step": 2651 + }, + { + "epoch": 3.137533274179237, + "grad_norm": 1.270472764968872, + "learning_rate": 1.6850000000000002e-06, + "loss": 0.0363, + "step": 2652 + }, + { + "epoch": 3.1387163561076603, + "grad_norm": 1.3482662439346313, + "learning_rate": 1.6837500000000002e-06, + "loss": 0.0355, + "step": 2653 + }, + { + "epoch": 3.139899438036084, + "grad_norm": 1.4808274507522583, + "learning_rate": 1.6825000000000002e-06, + "loss": 0.0328, + "step": 2654 + }, + { + "epoch": 3.1410825199645074, + "grad_norm": 1.4754095077514648, + "learning_rate": 1.6812500000000002e-06, + "loss": 0.0349, + "step": 2655 + }, + { + "epoch": 3.142265601892931, + "grad_norm": 1.7076677083969116, + "learning_rate": 1.6800000000000002e-06, + "loss": 0.0357, + "step": 2656 + }, + { + "epoch": 3.1434486838213545, + "grad_norm": 1.4881373643875122, + "learning_rate": 1.67875e-06, + "loss": 0.0362, + "step": 2657 + }, + { + "epoch": 3.144631765749778, + "grad_norm": 1.951103687286377, + "learning_rate": 1.6775000000000002e-06, + "loss": 0.0356, + "step": 2658 + }, + { + "epoch": 3.1458148476782015, + "grad_norm": 1.3693568706512451, + "learning_rate": 1.67625e-06, + "loss": 0.0315, + "step": 2659 + }, + { + "epoch": 3.1469979296066253, + "grad_norm": 1.3699461221694946, + "learning_rate": 1.6750000000000003e-06, + "loss": 0.0333, + "step": 2660 + }, + { + "epoch": 3.1481810115350486, + "grad_norm": 1.4908063411712646, + "learning_rate": 1.67375e-06, + "loss": 0.0339, + "step": 2661 + }, + { + "epoch": 3.1493640934634723, + "grad_norm": 1.4840742349624634, + "learning_rate": 1.6725000000000003e-06, + "loss": 0.0355, + "step": 2662 + }, + { + "epoch": 3.150547175391896, + "grad_norm": 1.2359851598739624, + "learning_rate": 1.67125e-06, + "loss": 0.0342, + "step": 2663 + }, + { + "epoch": 3.1517302573203194, + "grad_norm": 1.5540121793746948, + "learning_rate": 1.6700000000000003e-06, + "loss": 0.0372, + "step": 2664 + }, + { + "epoch": 3.152913339248743, + "grad_norm": 1.5254201889038086, + "learning_rate": 1.6687500000000001e-06, + "loss": 0.0344, + "step": 2665 + }, + { + "epoch": 3.1540964211771665, + "grad_norm": 1.4865429401397705, + "learning_rate": 1.6675000000000001e-06, + "loss": 0.043, + "step": 2666 + }, + { + "epoch": 3.1552795031055902, + "grad_norm": 1.3836324214935303, + "learning_rate": 1.6662500000000001e-06, + "loss": 0.0338, + "step": 2667 + }, + { + "epoch": 3.1564625850340136, + "grad_norm": 1.1533616781234741, + "learning_rate": 1.6650000000000002e-06, + "loss": 0.0322, + "step": 2668 + }, + { + "epoch": 3.1576456669624373, + "grad_norm": 1.3952782154083252, + "learning_rate": 1.6637500000000002e-06, + "loss": 0.0328, + "step": 2669 + }, + { + "epoch": 3.1588287488908606, + "grad_norm": 1.804916501045227, + "learning_rate": 1.6625000000000002e-06, + "loss": 0.0375, + "step": 2670 + }, + { + "epoch": 3.1600118308192844, + "grad_norm": 1.461647391319275, + "learning_rate": 1.6612500000000002e-06, + "loss": 0.0365, + "step": 2671 + }, + { + "epoch": 3.1611949127477077, + "grad_norm": 1.2532267570495605, + "learning_rate": 1.6600000000000002e-06, + "loss": 0.0322, + "step": 2672 + }, + { + "epoch": 3.1623779946761315, + "grad_norm": 1.2926342487335205, + "learning_rate": 1.65875e-06, + "loss": 0.0394, + "step": 2673 + }, + { + "epoch": 3.1635610766045548, + "grad_norm": 1.694517970085144, + "learning_rate": 1.6575000000000002e-06, + "loss": 0.0353, + "step": 2674 + }, + { + "epoch": 3.1647441585329785, + "grad_norm": 1.6156525611877441, + "learning_rate": 1.65625e-06, + "loss": 0.0362, + "step": 2675 + }, + { + "epoch": 3.165927240461402, + "grad_norm": 1.5529232025146484, + "learning_rate": 1.6550000000000002e-06, + "loss": 0.0369, + "step": 2676 + }, + { + "epoch": 3.1671103223898256, + "grad_norm": 1.2447835206985474, + "learning_rate": 1.65375e-06, + "loss": 0.0309, + "step": 2677 + }, + { + "epoch": 3.168293404318249, + "grad_norm": 1.426127552986145, + "learning_rate": 1.6525000000000003e-06, + "loss": 0.0358, + "step": 2678 + }, + { + "epoch": 3.1694764862466727, + "grad_norm": 1.5432547330856323, + "learning_rate": 1.65125e-06, + "loss": 0.0357, + "step": 2679 + }, + { + "epoch": 3.170659568175096, + "grad_norm": 1.5156960487365723, + "learning_rate": 1.6500000000000003e-06, + "loss": 0.0339, + "step": 2680 + }, + { + "epoch": 3.1718426501035197, + "grad_norm": 1.880261778831482, + "learning_rate": 1.64875e-06, + "loss": 0.0342, + "step": 2681 + }, + { + "epoch": 3.173025732031943, + "grad_norm": 1.4645737409591675, + "learning_rate": 1.6475000000000001e-06, + "loss": 0.0327, + "step": 2682 + }, + { + "epoch": 3.174208813960367, + "grad_norm": 1.5210113525390625, + "learning_rate": 1.6462500000000001e-06, + "loss": 0.0337, + "step": 2683 + }, + { + "epoch": 3.17539189588879, + "grad_norm": 1.3578792810440063, + "learning_rate": 1.6450000000000001e-06, + "loss": 0.0339, + "step": 2684 + }, + { + "epoch": 3.176574977817214, + "grad_norm": 1.3965708017349243, + "learning_rate": 1.6437500000000001e-06, + "loss": 0.0306, + "step": 2685 + }, + { + "epoch": 3.177758059745637, + "grad_norm": 1.466550350189209, + "learning_rate": 1.6425000000000002e-06, + "loss": 0.0316, + "step": 2686 + }, + { + "epoch": 3.178941141674061, + "grad_norm": 1.7264606952667236, + "learning_rate": 1.6412500000000002e-06, + "loss": 0.0346, + "step": 2687 + }, + { + "epoch": 3.1801242236024843, + "grad_norm": 1.4935555458068848, + "learning_rate": 1.6400000000000002e-06, + "loss": 0.0322, + "step": 2688 + }, + { + "epoch": 3.181307305530908, + "grad_norm": 1.392435908317566, + "learning_rate": 1.63875e-06, + "loss": 0.0311, + "step": 2689 + }, + { + "epoch": 3.1824903874593318, + "grad_norm": 1.4078010320663452, + "learning_rate": 1.6375000000000002e-06, + "loss": 0.0354, + "step": 2690 + }, + { + "epoch": 3.183673469387755, + "grad_norm": 1.585297703742981, + "learning_rate": 1.63625e-06, + "loss": 0.0365, + "step": 2691 + }, + { + "epoch": 3.184856551316179, + "grad_norm": 1.2957112789154053, + "learning_rate": 1.6350000000000002e-06, + "loss": 0.0331, + "step": 2692 + }, + { + "epoch": 3.186039633244602, + "grad_norm": 1.5087659358978271, + "learning_rate": 1.63375e-06, + "loss": 0.0353, + "step": 2693 + }, + { + "epoch": 3.187222715173026, + "grad_norm": 1.3302271366119385, + "learning_rate": 1.6325000000000003e-06, + "loss": 0.0376, + "step": 2694 + }, + { + "epoch": 3.1884057971014492, + "grad_norm": 1.5845961570739746, + "learning_rate": 1.63125e-06, + "loss": 0.0372, + "step": 2695 + }, + { + "epoch": 3.189588879029873, + "grad_norm": 1.5929687023162842, + "learning_rate": 1.6300000000000003e-06, + "loss": 0.037, + "step": 2696 + }, + { + "epoch": 3.1907719609582963, + "grad_norm": 1.6315062046051025, + "learning_rate": 1.62875e-06, + "loss": 0.0342, + "step": 2697 + }, + { + "epoch": 3.19195504288672, + "grad_norm": 1.526771903038025, + "learning_rate": 1.6275e-06, + "loss": 0.0329, + "step": 2698 + }, + { + "epoch": 3.1931381248151434, + "grad_norm": 1.7278059720993042, + "learning_rate": 1.62625e-06, + "loss": 0.0326, + "step": 2699 + }, + { + "epoch": 3.194321206743567, + "grad_norm": 1.126779556274414, + "learning_rate": 1.6250000000000001e-06, + "loss": 0.0341, + "step": 2700 + }, + { + "epoch": 3.1955042886719904, + "grad_norm": 1.5368411540985107, + "learning_rate": 1.6237500000000001e-06, + "loss": 0.0374, + "step": 2701 + }, + { + "epoch": 3.196687370600414, + "grad_norm": 1.3360999822616577, + "learning_rate": 1.6225000000000001e-06, + "loss": 0.0345, + "step": 2702 + }, + { + "epoch": 3.1978704525288375, + "grad_norm": 1.6106908321380615, + "learning_rate": 1.6212500000000001e-06, + "loss": 0.0378, + "step": 2703 + }, + { + "epoch": 3.1990535344572613, + "grad_norm": 1.366951584815979, + "learning_rate": 1.6200000000000002e-06, + "loss": 0.0366, + "step": 2704 + }, + { + "epoch": 3.2002366163856846, + "grad_norm": 1.4330976009368896, + "learning_rate": 1.61875e-06, + "loss": 0.0331, + "step": 2705 + }, + { + "epoch": 3.2014196983141083, + "grad_norm": 1.664880633354187, + "learning_rate": 1.6175000000000002e-06, + "loss": 0.0309, + "step": 2706 + }, + { + "epoch": 3.2026027802425316, + "grad_norm": 1.5748622417449951, + "learning_rate": 1.61625e-06, + "loss": 0.0333, + "step": 2707 + }, + { + "epoch": 3.2037858621709554, + "grad_norm": 1.4909648895263672, + "learning_rate": 1.6150000000000002e-06, + "loss": 0.0352, + "step": 2708 + }, + { + "epoch": 3.2049689440993787, + "grad_norm": 1.4156098365783691, + "learning_rate": 1.61375e-06, + "loss": 0.0338, + "step": 2709 + }, + { + "epoch": 3.2061520260278025, + "grad_norm": 1.4368846416473389, + "learning_rate": 1.6125000000000002e-06, + "loss": 0.0351, + "step": 2710 + }, + { + "epoch": 3.207335107956226, + "grad_norm": 1.7556521892547607, + "learning_rate": 1.61125e-06, + "loss": 0.0315, + "step": 2711 + }, + { + "epoch": 3.2085181898846495, + "grad_norm": 1.416122317314148, + "learning_rate": 1.6100000000000003e-06, + "loss": 0.0378, + "step": 2712 + }, + { + "epoch": 3.209701271813073, + "grad_norm": 1.3234834671020508, + "learning_rate": 1.60875e-06, + "loss": 0.0358, + "step": 2713 + }, + { + "epoch": 3.2108843537414966, + "grad_norm": 1.345319390296936, + "learning_rate": 1.6075e-06, + "loss": 0.039, + "step": 2714 + }, + { + "epoch": 3.21206743566992, + "grad_norm": 1.5577281713485718, + "learning_rate": 1.60625e-06, + "loss": 0.0358, + "step": 2715 + }, + { + "epoch": 3.2132505175983437, + "grad_norm": 1.6010186672210693, + "learning_rate": 1.605e-06, + "loss": 0.0348, + "step": 2716 + }, + { + "epoch": 3.2144335995267674, + "grad_norm": 1.2769134044647217, + "learning_rate": 1.60375e-06, + "loss": 0.0396, + "step": 2717 + }, + { + "epoch": 3.2156166814551908, + "grad_norm": 1.500512719154358, + "learning_rate": 1.6025000000000001e-06, + "loss": 0.0331, + "step": 2718 + }, + { + "epoch": 3.2167997633836145, + "grad_norm": 1.5311400890350342, + "learning_rate": 1.6012500000000001e-06, + "loss": 0.0357, + "step": 2719 + }, + { + "epoch": 3.217982845312038, + "grad_norm": 1.2612730264663696, + "learning_rate": 1.6000000000000001e-06, + "loss": 0.0319, + "step": 2720 + }, + { + "epoch": 3.2191659272404616, + "grad_norm": 1.4925894737243652, + "learning_rate": 1.59875e-06, + "loss": 0.0301, + "step": 2721 + }, + { + "epoch": 3.220349009168885, + "grad_norm": 1.8070454597473145, + "learning_rate": 1.5975000000000002e-06, + "loss": 0.0354, + "step": 2722 + }, + { + "epoch": 3.2215320910973086, + "grad_norm": 1.3110291957855225, + "learning_rate": 1.59625e-06, + "loss": 0.0346, + "step": 2723 + }, + { + "epoch": 3.222715173025732, + "grad_norm": 1.7656325101852417, + "learning_rate": 1.5950000000000002e-06, + "loss": 0.0378, + "step": 2724 + }, + { + "epoch": 3.2238982549541557, + "grad_norm": 1.501821517944336, + "learning_rate": 1.59375e-06, + "loss": 0.0324, + "step": 2725 + }, + { + "epoch": 3.225081336882579, + "grad_norm": 1.57681143283844, + "learning_rate": 1.5925000000000002e-06, + "loss": 0.0342, + "step": 2726 + }, + { + "epoch": 3.226264418811003, + "grad_norm": 1.1805040836334229, + "learning_rate": 1.59125e-06, + "loss": 0.0314, + "step": 2727 + }, + { + "epoch": 3.227447500739426, + "grad_norm": 1.7549033164978027, + "learning_rate": 1.5900000000000002e-06, + "loss": 0.0309, + "step": 2728 + }, + { + "epoch": 3.22863058266785, + "grad_norm": 1.5704357624053955, + "learning_rate": 1.58875e-06, + "loss": 0.0381, + "step": 2729 + }, + { + "epoch": 3.229813664596273, + "grad_norm": 1.302894949913025, + "learning_rate": 1.5875e-06, + "loss": 0.0365, + "step": 2730 + }, + { + "epoch": 3.230996746524697, + "grad_norm": 1.2623999118804932, + "learning_rate": 1.58625e-06, + "loss": 0.0344, + "step": 2731 + }, + { + "epoch": 3.2321798284531202, + "grad_norm": 1.4486894607543945, + "learning_rate": 1.585e-06, + "loss": 0.0345, + "step": 2732 + }, + { + "epoch": 3.233362910381544, + "grad_norm": 1.318286418914795, + "learning_rate": 1.58375e-06, + "loss": 0.0314, + "step": 2733 + }, + { + "epoch": 3.2345459923099673, + "grad_norm": 1.30774986743927, + "learning_rate": 1.5825e-06, + "loss": 0.0343, + "step": 2734 + }, + { + "epoch": 3.235729074238391, + "grad_norm": 1.4238911867141724, + "learning_rate": 1.5812500000000001e-06, + "loss": 0.0351, + "step": 2735 + }, + { + "epoch": 3.2369121561668144, + "grad_norm": 1.4870507717132568, + "learning_rate": 1.5800000000000001e-06, + "loss": 0.0326, + "step": 2736 + }, + { + "epoch": 3.238095238095238, + "grad_norm": 1.3641709089279175, + "learning_rate": 1.57875e-06, + "loss": 0.035, + "step": 2737 + }, + { + "epoch": 3.2392783200236615, + "grad_norm": 1.5676878690719604, + "learning_rate": 1.5775000000000001e-06, + "loss": 0.0362, + "step": 2738 + }, + { + "epoch": 3.240461401952085, + "grad_norm": 1.366088628768921, + "learning_rate": 1.57625e-06, + "loss": 0.0302, + "step": 2739 + }, + { + "epoch": 3.2416444838805085, + "grad_norm": 1.4484782218933105, + "learning_rate": 1.5750000000000002e-06, + "loss": 0.0367, + "step": 2740 + }, + { + "epoch": 3.2428275658089323, + "grad_norm": 1.5502800941467285, + "learning_rate": 1.57375e-06, + "loss": 0.0351, + "step": 2741 + }, + { + "epoch": 3.2440106477373556, + "grad_norm": 1.425485610961914, + "learning_rate": 1.5725000000000002e-06, + "loss": 0.0382, + "step": 2742 + }, + { + "epoch": 3.2451937296657793, + "grad_norm": 1.6285887956619263, + "learning_rate": 1.57125e-06, + "loss": 0.0322, + "step": 2743 + }, + { + "epoch": 3.246376811594203, + "grad_norm": 1.7583813667297363, + "learning_rate": 1.5700000000000002e-06, + "loss": 0.0332, + "step": 2744 + }, + { + "epoch": 3.2475598935226264, + "grad_norm": 1.3697471618652344, + "learning_rate": 1.56875e-06, + "loss": 0.0318, + "step": 2745 + }, + { + "epoch": 3.24874297545105, + "grad_norm": 1.433937430381775, + "learning_rate": 1.5675e-06, + "loss": 0.0336, + "step": 2746 + }, + { + "epoch": 3.2499260573794735, + "grad_norm": 1.3876763582229614, + "learning_rate": 1.56625e-06, + "loss": 0.0329, + "step": 2747 + }, + { + "epoch": 3.2511091393078972, + "grad_norm": 1.3260279893875122, + "learning_rate": 1.565e-06, + "loss": 0.0337, + "step": 2748 + }, + { + "epoch": 3.2522922212363206, + "grad_norm": 1.2209817171096802, + "learning_rate": 1.56375e-06, + "loss": 0.0316, + "step": 2749 + }, + { + "epoch": 3.2534753031647443, + "grad_norm": 1.6898490190505981, + "learning_rate": 1.5625e-06, + "loss": 0.0403, + "step": 2750 + }, + { + "epoch": 3.2546583850931676, + "grad_norm": 1.3076483011245728, + "learning_rate": 1.5612500000000003e-06, + "loss": 0.0312, + "step": 2751 + }, + { + "epoch": 3.2558414670215914, + "grad_norm": 1.7955926656723022, + "learning_rate": 1.56e-06, + "loss": 0.034, + "step": 2752 + }, + { + "epoch": 3.2570245489500147, + "grad_norm": 1.7558958530426025, + "learning_rate": 1.5587500000000003e-06, + "loss": 0.0359, + "step": 2753 + }, + { + "epoch": 3.2582076308784385, + "grad_norm": 1.3676015138626099, + "learning_rate": 1.5575000000000001e-06, + "loss": 0.0315, + "step": 2754 + }, + { + "epoch": 3.2593907128068618, + "grad_norm": 1.3773105144500732, + "learning_rate": 1.5562500000000003e-06, + "loss": 0.0348, + "step": 2755 + }, + { + "epoch": 3.2605737947352855, + "grad_norm": 1.4999548196792603, + "learning_rate": 1.5550000000000001e-06, + "loss": 0.0304, + "step": 2756 + }, + { + "epoch": 3.261756876663709, + "grad_norm": 1.65998113155365, + "learning_rate": 1.5537500000000002e-06, + "loss": 0.0398, + "step": 2757 + }, + { + "epoch": 3.2629399585921326, + "grad_norm": 1.6981534957885742, + "learning_rate": 1.5525000000000002e-06, + "loss": 0.0334, + "step": 2758 + }, + { + "epoch": 3.264123040520556, + "grad_norm": 1.7219655513763428, + "learning_rate": 1.5512500000000002e-06, + "loss": 0.0382, + "step": 2759 + }, + { + "epoch": 3.2653061224489797, + "grad_norm": 1.7153385877609253, + "learning_rate": 1.5500000000000002e-06, + "loss": 0.0382, + "step": 2760 + }, + { + "epoch": 3.266489204377403, + "grad_norm": 1.5698245763778687, + "learning_rate": 1.5487500000000002e-06, + "loss": 0.0378, + "step": 2761 + }, + { + "epoch": 3.2676722863058267, + "grad_norm": 1.7160028219223022, + "learning_rate": 1.5475000000000002e-06, + "loss": 0.0395, + "step": 2762 + }, + { + "epoch": 3.26885536823425, + "grad_norm": 1.5031168460845947, + "learning_rate": 1.5462500000000002e-06, + "loss": 0.0388, + "step": 2763 + }, + { + "epoch": 3.270038450162674, + "grad_norm": 1.401690125465393, + "learning_rate": 1.545e-06, + "loss": 0.033, + "step": 2764 + }, + { + "epoch": 3.271221532091097, + "grad_norm": 1.4632117748260498, + "learning_rate": 1.5437500000000003e-06, + "loss": 0.0333, + "step": 2765 + }, + { + "epoch": 3.272404614019521, + "grad_norm": 1.7546350955963135, + "learning_rate": 1.5425e-06, + "loss": 0.0358, + "step": 2766 + }, + { + "epoch": 3.2735876959479446, + "grad_norm": 1.6384639739990234, + "learning_rate": 1.5412500000000003e-06, + "loss": 0.0326, + "step": 2767 + }, + { + "epoch": 3.274770777876368, + "grad_norm": 1.6297237873077393, + "learning_rate": 1.54e-06, + "loss": 0.0429, + "step": 2768 + }, + { + "epoch": 3.2759538598047913, + "grad_norm": 1.4283443689346313, + "learning_rate": 1.5387500000000003e-06, + "loss": 0.0308, + "step": 2769 + }, + { + "epoch": 3.277136941733215, + "grad_norm": 1.428305745124817, + "learning_rate": 1.5375e-06, + "loss": 0.038, + "step": 2770 + }, + { + "epoch": 3.2783200236616388, + "grad_norm": 1.7066211700439453, + "learning_rate": 1.5362500000000003e-06, + "loss": 0.032, + "step": 2771 + }, + { + "epoch": 3.279503105590062, + "grad_norm": 1.3155156373977661, + "learning_rate": 1.5350000000000001e-06, + "loss": 0.0308, + "step": 2772 + }, + { + "epoch": 3.280686187518486, + "grad_norm": 1.3547600507736206, + "learning_rate": 1.5337500000000001e-06, + "loss": 0.0327, + "step": 2773 + }, + { + "epoch": 3.281869269446909, + "grad_norm": 1.5091322660446167, + "learning_rate": 1.5325000000000002e-06, + "loss": 0.0373, + "step": 2774 + }, + { + "epoch": 3.283052351375333, + "grad_norm": 1.4633378982543945, + "learning_rate": 1.5312500000000002e-06, + "loss": 0.041, + "step": 2775 + }, + { + "epoch": 3.2842354333037562, + "grad_norm": 1.3562734127044678, + "learning_rate": 1.5300000000000002e-06, + "loss": 0.0325, + "step": 2776 + }, + { + "epoch": 3.28541851523218, + "grad_norm": 1.602185606956482, + "learning_rate": 1.5287500000000002e-06, + "loss": 0.0316, + "step": 2777 + }, + { + "epoch": 3.2866015971606033, + "grad_norm": 1.4459859132766724, + "learning_rate": 1.5275000000000002e-06, + "loss": 0.0351, + "step": 2778 + }, + { + "epoch": 3.287784679089027, + "grad_norm": 1.5817079544067383, + "learning_rate": 1.5262500000000002e-06, + "loss": 0.037, + "step": 2779 + }, + { + "epoch": 3.2889677610174504, + "grad_norm": 1.5800535678863525, + "learning_rate": 1.525e-06, + "loss": 0.0312, + "step": 2780 + }, + { + "epoch": 3.290150842945874, + "grad_norm": 1.6160582304000854, + "learning_rate": 1.5237500000000002e-06, + "loss": 0.0313, + "step": 2781 + }, + { + "epoch": 3.2913339248742974, + "grad_norm": 1.3111237287521362, + "learning_rate": 1.5225e-06, + "loss": 0.0329, + "step": 2782 + }, + { + "epoch": 3.292517006802721, + "grad_norm": 1.543649435043335, + "learning_rate": 1.5212500000000003e-06, + "loss": 0.0336, + "step": 2783 + }, + { + "epoch": 3.2937000887311445, + "grad_norm": 1.7419538497924805, + "learning_rate": 1.52e-06, + "loss": 0.0375, + "step": 2784 + }, + { + "epoch": 3.2948831706595683, + "grad_norm": 1.4487885236740112, + "learning_rate": 1.5187500000000003e-06, + "loss": 0.0341, + "step": 2785 + }, + { + "epoch": 3.2960662525879916, + "grad_norm": 1.4753068685531616, + "learning_rate": 1.5175e-06, + "loss": 0.0327, + "step": 2786 + }, + { + "epoch": 3.2972493345164153, + "grad_norm": 1.6539093255996704, + "learning_rate": 1.5162500000000003e-06, + "loss": 0.0341, + "step": 2787 + }, + { + "epoch": 3.2984324164448386, + "grad_norm": 1.633671522140503, + "learning_rate": 1.5150000000000001e-06, + "loss": 0.0352, + "step": 2788 + }, + { + "epoch": 3.2996154983732624, + "grad_norm": 1.473493218421936, + "learning_rate": 1.5137500000000001e-06, + "loss": 0.035, + "step": 2789 + }, + { + "epoch": 3.3007985803016857, + "grad_norm": 1.343234896659851, + "learning_rate": 1.5125000000000001e-06, + "loss": 0.0288, + "step": 2790 + }, + { + "epoch": 3.3019816622301095, + "grad_norm": 1.4213192462921143, + "learning_rate": 1.5112500000000001e-06, + "loss": 0.033, + "step": 2791 + }, + { + "epoch": 3.303164744158533, + "grad_norm": 1.4502743482589722, + "learning_rate": 1.5100000000000002e-06, + "loss": 0.0298, + "step": 2792 + }, + { + "epoch": 3.3043478260869565, + "grad_norm": 1.7088652849197388, + "learning_rate": 1.5087500000000002e-06, + "loss": 0.0403, + "step": 2793 + }, + { + "epoch": 3.3055309080153803, + "grad_norm": 1.388545036315918, + "learning_rate": 1.5075000000000002e-06, + "loss": 0.0326, + "step": 2794 + }, + { + "epoch": 3.3067139899438036, + "grad_norm": 1.2549660205841064, + "learning_rate": 1.5062500000000002e-06, + "loss": 0.0346, + "step": 2795 + }, + { + "epoch": 3.307897071872227, + "grad_norm": 1.3282636404037476, + "learning_rate": 1.505e-06, + "loss": 0.0347, + "step": 2796 + }, + { + "epoch": 3.3090801538006507, + "grad_norm": 1.6094398498535156, + "learning_rate": 1.5037500000000002e-06, + "loss": 0.0421, + "step": 2797 + }, + { + "epoch": 3.3102632357290744, + "grad_norm": 1.650476336479187, + "learning_rate": 1.5025e-06, + "loss": 0.0306, + "step": 2798 + }, + { + "epoch": 3.3114463176574978, + "grad_norm": 1.3900727033615112, + "learning_rate": 1.5012500000000002e-06, + "loss": 0.0356, + "step": 2799 + }, + { + "epoch": 3.3126293995859215, + "grad_norm": 1.6523176431655884, + "learning_rate": 1.5e-06, + "loss": 0.0369, + "step": 2800 + }, + { + "epoch": 3.313812481514345, + "grad_norm": 1.6868435144424438, + "learning_rate": 1.4987500000000003e-06, + "loss": 0.0336, + "step": 2801 + }, + { + "epoch": 3.3149955634427686, + "grad_norm": 1.5165168046951294, + "learning_rate": 1.4975e-06, + "loss": 0.0319, + "step": 2802 + }, + { + "epoch": 3.316178645371192, + "grad_norm": 1.4603636264801025, + "learning_rate": 1.4962500000000003e-06, + "loss": 0.0289, + "step": 2803 + }, + { + "epoch": 3.3173617272996156, + "grad_norm": 1.3170220851898193, + "learning_rate": 1.495e-06, + "loss": 0.0295, + "step": 2804 + }, + { + "epoch": 3.318544809228039, + "grad_norm": 3.3199410438537598, + "learning_rate": 1.49375e-06, + "loss": 0.0286, + "step": 2805 + }, + { + "epoch": 3.3197278911564627, + "grad_norm": 1.7035584449768066, + "learning_rate": 1.4925000000000001e-06, + "loss": 0.0329, + "step": 2806 + }, + { + "epoch": 3.320910973084886, + "grad_norm": 1.4776946306228638, + "learning_rate": 1.4912500000000001e-06, + "loss": 0.0368, + "step": 2807 + }, + { + "epoch": 3.32209405501331, + "grad_norm": 1.7139945030212402, + "learning_rate": 1.4900000000000001e-06, + "loss": 0.0298, + "step": 2808 + }, + { + "epoch": 3.323277136941733, + "grad_norm": 1.5534343719482422, + "learning_rate": 1.4887500000000001e-06, + "loss": 0.0381, + "step": 2809 + }, + { + "epoch": 3.324460218870157, + "grad_norm": 1.4955379962921143, + "learning_rate": 1.4875000000000002e-06, + "loss": 0.0345, + "step": 2810 + }, + { + "epoch": 3.32564330079858, + "grad_norm": 1.404374599456787, + "learning_rate": 1.4862500000000002e-06, + "loss": 0.0329, + "step": 2811 + }, + { + "epoch": 3.326826382727004, + "grad_norm": 1.5724259614944458, + "learning_rate": 1.485e-06, + "loss": 0.0406, + "step": 2812 + }, + { + "epoch": 3.3280094646554272, + "grad_norm": 1.4024325609207153, + "learning_rate": 1.4837500000000002e-06, + "loss": 0.0349, + "step": 2813 + }, + { + "epoch": 3.329192546583851, + "grad_norm": 1.460828185081482, + "learning_rate": 1.4825e-06, + "loss": 0.0346, + "step": 2814 + }, + { + "epoch": 3.3303756285122743, + "grad_norm": 1.4691442251205444, + "learning_rate": 1.4812500000000002e-06, + "loss": 0.0355, + "step": 2815 + }, + { + "epoch": 3.331558710440698, + "grad_norm": 1.6167815923690796, + "learning_rate": 1.48e-06, + "loss": 0.0379, + "step": 2816 + }, + { + "epoch": 3.3327417923691214, + "grad_norm": 1.7537893056869507, + "learning_rate": 1.4787500000000002e-06, + "loss": 0.0395, + "step": 2817 + }, + { + "epoch": 3.333924874297545, + "grad_norm": 1.255554437637329, + "learning_rate": 1.4775e-06, + "loss": 0.0336, + "step": 2818 + }, + { + "epoch": 3.3351079562259685, + "grad_norm": 1.3924132585525513, + "learning_rate": 1.4762500000000003e-06, + "loss": 0.0326, + "step": 2819 + }, + { + "epoch": 3.336291038154392, + "grad_norm": 1.4285389184951782, + "learning_rate": 1.475e-06, + "loss": 0.0373, + "step": 2820 + }, + { + "epoch": 3.337474120082816, + "grad_norm": 1.338178277015686, + "learning_rate": 1.47375e-06, + "loss": 0.0372, + "step": 2821 + }, + { + "epoch": 3.3386572020112393, + "grad_norm": 1.7925024032592773, + "learning_rate": 1.4725e-06, + "loss": 0.0382, + "step": 2822 + }, + { + "epoch": 3.3398402839396626, + "grad_norm": 1.488305687904358, + "learning_rate": 1.47125e-06, + "loss": 0.0335, + "step": 2823 + }, + { + "epoch": 3.3410233658680863, + "grad_norm": 1.1806750297546387, + "learning_rate": 1.4700000000000001e-06, + "loss": 0.0377, + "step": 2824 + }, + { + "epoch": 3.34220644779651, + "grad_norm": 1.3681340217590332, + "learning_rate": 1.4687500000000001e-06, + "loss": 0.0325, + "step": 2825 + }, + { + "epoch": 3.3433895297249334, + "grad_norm": 1.4270672798156738, + "learning_rate": 1.4675000000000001e-06, + "loss": 0.0365, + "step": 2826 + }, + { + "epoch": 3.344572611653357, + "grad_norm": 1.5541542768478394, + "learning_rate": 1.4662500000000001e-06, + "loss": 0.0312, + "step": 2827 + }, + { + "epoch": 3.3457556935817805, + "grad_norm": 1.6314609050750732, + "learning_rate": 1.465e-06, + "loss": 0.0379, + "step": 2828 + }, + { + "epoch": 3.3469387755102042, + "grad_norm": 1.501426100730896, + "learning_rate": 1.4637500000000002e-06, + "loss": 0.0338, + "step": 2829 + }, + { + "epoch": 3.3481218574386276, + "grad_norm": 1.356476902961731, + "learning_rate": 1.4625e-06, + "loss": 0.0298, + "step": 2830 + }, + { + "epoch": 3.3493049393670513, + "grad_norm": 1.5355725288391113, + "learning_rate": 1.4612500000000002e-06, + "loss": 0.0377, + "step": 2831 + }, + { + "epoch": 3.3504880212954746, + "grad_norm": 1.5467458963394165, + "learning_rate": 1.46e-06, + "loss": 0.0337, + "step": 2832 + }, + { + "epoch": 3.3516711032238984, + "grad_norm": 1.5599170923233032, + "learning_rate": 1.4587500000000002e-06, + "loss": 0.0349, + "step": 2833 + }, + { + "epoch": 3.3528541851523217, + "grad_norm": 1.2871712446212769, + "learning_rate": 1.4575e-06, + "loss": 0.032, + "step": 2834 + }, + { + "epoch": 3.3540372670807455, + "grad_norm": 1.3052334785461426, + "learning_rate": 1.4562500000000002e-06, + "loss": 0.0304, + "step": 2835 + }, + { + "epoch": 3.3552203490091688, + "grad_norm": 1.0657607316970825, + "learning_rate": 1.455e-06, + "loss": 0.0315, + "step": 2836 + }, + { + "epoch": 3.3564034309375925, + "grad_norm": 1.5395492315292358, + "learning_rate": 1.45375e-06, + "loss": 0.0335, + "step": 2837 + }, + { + "epoch": 3.357586512866016, + "grad_norm": 1.5174024105072021, + "learning_rate": 1.4525e-06, + "loss": 0.0341, + "step": 2838 + }, + { + "epoch": 3.3587695947944396, + "grad_norm": 1.5007987022399902, + "learning_rate": 1.45125e-06, + "loss": 0.0413, + "step": 2839 + }, + { + "epoch": 3.359952676722863, + "grad_norm": 1.1538548469543457, + "learning_rate": 1.45e-06, + "loss": 0.0312, + "step": 2840 + }, + { + "epoch": 3.3611357586512867, + "grad_norm": 1.4360628128051758, + "learning_rate": 1.44875e-06, + "loss": 0.0386, + "step": 2841 + }, + { + "epoch": 3.36231884057971, + "grad_norm": 1.3586925268173218, + "learning_rate": 1.4475000000000001e-06, + "loss": 0.0379, + "step": 2842 + }, + { + "epoch": 3.3635019225081337, + "grad_norm": 1.4885003566741943, + "learning_rate": 1.4462500000000001e-06, + "loss": 0.0342, + "step": 2843 + }, + { + "epoch": 3.364685004436557, + "grad_norm": 1.4755313396453857, + "learning_rate": 1.445e-06, + "loss": 0.0366, + "step": 2844 + }, + { + "epoch": 3.365868086364981, + "grad_norm": 1.4937607049942017, + "learning_rate": 1.4437500000000002e-06, + "loss": 0.0296, + "step": 2845 + }, + { + "epoch": 3.367051168293404, + "grad_norm": 1.3136570453643799, + "learning_rate": 1.4425e-06, + "loss": 0.0331, + "step": 2846 + }, + { + "epoch": 3.368234250221828, + "grad_norm": 1.3184014558792114, + "learning_rate": 1.4412500000000002e-06, + "loss": 0.0296, + "step": 2847 + }, + { + "epoch": 3.3694173321502516, + "grad_norm": 1.6608057022094727, + "learning_rate": 1.44e-06, + "loss": 0.033, + "step": 2848 + }, + { + "epoch": 3.370600414078675, + "grad_norm": 1.7398295402526855, + "learning_rate": 1.4387500000000002e-06, + "loss": 0.0367, + "step": 2849 + }, + { + "epoch": 3.3717834960070983, + "grad_norm": 1.5712093114852905, + "learning_rate": 1.4375e-06, + "loss": 0.0346, + "step": 2850 + }, + { + "epoch": 3.372966577935522, + "grad_norm": 2.189771890640259, + "learning_rate": 1.4362500000000002e-06, + "loss": 0.0386, + "step": 2851 + }, + { + "epoch": 3.3741496598639458, + "grad_norm": 1.3624439239501953, + "learning_rate": 1.435e-06, + "loss": 0.0358, + "step": 2852 + }, + { + "epoch": 3.375332741792369, + "grad_norm": 1.8808847665786743, + "learning_rate": 1.43375e-06, + "loss": 0.0375, + "step": 2853 + }, + { + "epoch": 3.376515823720793, + "grad_norm": 1.358487844467163, + "learning_rate": 1.4325e-06, + "loss": 0.037, + "step": 2854 + }, + { + "epoch": 3.377698905649216, + "grad_norm": 1.4706295728683472, + "learning_rate": 1.43125e-06, + "loss": 0.0341, + "step": 2855 + }, + { + "epoch": 3.37888198757764, + "grad_norm": 1.548400640487671, + "learning_rate": 1.43e-06, + "loss": 0.0337, + "step": 2856 + }, + { + "epoch": 3.3800650695060632, + "grad_norm": 1.8608593940734863, + "learning_rate": 1.42875e-06, + "loss": 0.0352, + "step": 2857 + }, + { + "epoch": 3.381248151434487, + "grad_norm": 1.1719350814819336, + "learning_rate": 1.4275e-06, + "loss": 0.0339, + "step": 2858 + }, + { + "epoch": 3.3824312333629103, + "grad_norm": 1.5975359678268433, + "learning_rate": 1.4262500000000001e-06, + "loss": 0.0345, + "step": 2859 + }, + { + "epoch": 3.383614315291334, + "grad_norm": 1.7109965085983276, + "learning_rate": 1.425e-06, + "loss": 0.0363, + "step": 2860 + }, + { + "epoch": 3.3847973972197574, + "grad_norm": 1.6883807182312012, + "learning_rate": 1.4237500000000001e-06, + "loss": 0.0341, + "step": 2861 + }, + { + "epoch": 3.385980479148181, + "grad_norm": 1.166603684425354, + "learning_rate": 1.4225e-06, + "loss": 0.0332, + "step": 2862 + }, + { + "epoch": 3.3871635610766044, + "grad_norm": 1.5817257165908813, + "learning_rate": 1.4212500000000002e-06, + "loss": 0.0365, + "step": 2863 + }, + { + "epoch": 3.388346643005028, + "grad_norm": 1.8719844818115234, + "learning_rate": 1.42e-06, + "loss": 0.0341, + "step": 2864 + }, + { + "epoch": 3.3895297249334515, + "grad_norm": 1.7708829641342163, + "learning_rate": 1.4187500000000002e-06, + "loss": 0.0395, + "step": 2865 + }, + { + "epoch": 3.3907128068618753, + "grad_norm": 1.7726995944976807, + "learning_rate": 1.4175e-06, + "loss": 0.0356, + "step": 2866 + }, + { + "epoch": 3.3918958887902986, + "grad_norm": 1.3731539249420166, + "learning_rate": 1.4162500000000002e-06, + "loss": 0.0446, + "step": 2867 + }, + { + "epoch": 3.3930789707187223, + "grad_norm": 1.4510589838027954, + "learning_rate": 1.415e-06, + "loss": 0.0393, + "step": 2868 + }, + { + "epoch": 3.3942620526471456, + "grad_norm": 1.5432417392730713, + "learning_rate": 1.41375e-06, + "loss": 0.038, + "step": 2869 + }, + { + "epoch": 3.3954451345755694, + "grad_norm": 1.2934036254882812, + "learning_rate": 1.4125e-06, + "loss": 0.034, + "step": 2870 + }, + { + "epoch": 3.3966282165039927, + "grad_norm": 1.3716113567352295, + "learning_rate": 1.41125e-06, + "loss": 0.036, + "step": 2871 + }, + { + "epoch": 3.3978112984324165, + "grad_norm": 1.6273936033248901, + "learning_rate": 1.41e-06, + "loss": 0.0339, + "step": 2872 + }, + { + "epoch": 3.39899438036084, + "grad_norm": 1.334596872329712, + "learning_rate": 1.40875e-06, + "loss": 0.0306, + "step": 2873 + }, + { + "epoch": 3.4001774622892635, + "grad_norm": 1.4858254194259644, + "learning_rate": 1.4075e-06, + "loss": 0.0349, + "step": 2874 + }, + { + "epoch": 3.4013605442176873, + "grad_norm": 1.631413221359253, + "learning_rate": 1.40625e-06, + "loss": 0.0357, + "step": 2875 + }, + { + "epoch": 3.4025436261461106, + "grad_norm": 1.8111709356307983, + "learning_rate": 1.4050000000000003e-06, + "loss": 0.0325, + "step": 2876 + }, + { + "epoch": 3.403726708074534, + "grad_norm": 1.7803117036819458, + "learning_rate": 1.4037500000000001e-06, + "loss": 0.0363, + "step": 2877 + }, + { + "epoch": 3.4049097900029577, + "grad_norm": 1.5928584337234497, + "learning_rate": 1.4025000000000003e-06, + "loss": 0.0354, + "step": 2878 + }, + { + "epoch": 3.4060928719313814, + "grad_norm": 1.471239447593689, + "learning_rate": 1.4012500000000001e-06, + "loss": 0.0309, + "step": 2879 + }, + { + "epoch": 3.4072759538598048, + "grad_norm": 1.6542584896087646, + "learning_rate": 1.4000000000000001e-06, + "loss": 0.0342, + "step": 2880 + }, + { + "epoch": 3.4084590357882285, + "grad_norm": 1.242817997932434, + "learning_rate": 1.3987500000000002e-06, + "loss": 0.0329, + "step": 2881 + }, + { + "epoch": 3.409642117716652, + "grad_norm": 1.6571860313415527, + "learning_rate": 1.3975000000000002e-06, + "loss": 0.0353, + "step": 2882 + }, + { + "epoch": 3.4108251996450756, + "grad_norm": 1.554930567741394, + "learning_rate": 1.3962500000000002e-06, + "loss": 0.036, + "step": 2883 + }, + { + "epoch": 3.412008281573499, + "grad_norm": 1.3862043619155884, + "learning_rate": 1.3950000000000002e-06, + "loss": 0.0373, + "step": 2884 + }, + { + "epoch": 3.4131913635019226, + "grad_norm": 1.3036785125732422, + "learning_rate": 1.3937500000000002e-06, + "loss": 0.0331, + "step": 2885 + }, + { + "epoch": 3.414374445430346, + "grad_norm": 1.5559929609298706, + "learning_rate": 1.3925000000000002e-06, + "loss": 0.0335, + "step": 2886 + }, + { + "epoch": 3.4155575273587697, + "grad_norm": 1.4783521890640259, + "learning_rate": 1.39125e-06, + "loss": 0.032, + "step": 2887 + }, + { + "epoch": 3.416740609287193, + "grad_norm": 1.4191184043884277, + "learning_rate": 1.3900000000000002e-06, + "loss": 0.034, + "step": 2888 + }, + { + "epoch": 3.417923691215617, + "grad_norm": 1.601242184638977, + "learning_rate": 1.38875e-06, + "loss": 0.0331, + "step": 2889 + }, + { + "epoch": 3.41910677314404, + "grad_norm": 1.8074856996536255, + "learning_rate": 1.3875000000000003e-06, + "loss": 0.0381, + "step": 2890 + }, + { + "epoch": 3.420289855072464, + "grad_norm": 1.66084623336792, + "learning_rate": 1.38625e-06, + "loss": 0.0351, + "step": 2891 + }, + { + "epoch": 3.421472937000887, + "grad_norm": 2.1125540733337402, + "learning_rate": 1.3850000000000003e-06, + "loss": 0.0368, + "step": 2892 + }, + { + "epoch": 3.422656018929311, + "grad_norm": 1.7980034351348877, + "learning_rate": 1.38375e-06, + "loss": 0.0373, + "step": 2893 + }, + { + "epoch": 3.4238391008577342, + "grad_norm": 1.2917977571487427, + "learning_rate": 1.3825000000000003e-06, + "loss": 0.033, + "step": 2894 + }, + { + "epoch": 3.425022182786158, + "grad_norm": 1.5396780967712402, + "learning_rate": 1.3812500000000001e-06, + "loss": 0.0325, + "step": 2895 + }, + { + "epoch": 3.4262052647145813, + "grad_norm": 1.5898183584213257, + "learning_rate": 1.3800000000000001e-06, + "loss": 0.0358, + "step": 2896 + }, + { + "epoch": 3.427388346643005, + "grad_norm": 1.4854717254638672, + "learning_rate": 1.3787500000000001e-06, + "loss": 0.033, + "step": 2897 + }, + { + "epoch": 3.4285714285714284, + "grad_norm": 1.4945628643035889, + "learning_rate": 1.3775000000000002e-06, + "loss": 0.0335, + "step": 2898 + }, + { + "epoch": 3.429754510499852, + "grad_norm": 1.7546871900558472, + "learning_rate": 1.3762500000000002e-06, + "loss": 0.0355, + "step": 2899 + }, + { + "epoch": 3.4309375924282755, + "grad_norm": 1.5036351680755615, + "learning_rate": 1.3750000000000002e-06, + "loss": 0.0307, + "step": 2900 + }, + { + "epoch": 3.432120674356699, + "grad_norm": 1.3347246646881104, + "learning_rate": 1.3737500000000002e-06, + "loss": 0.0328, + "step": 2901 + }, + { + "epoch": 3.433303756285123, + "grad_norm": 1.319620966911316, + "learning_rate": 1.3725000000000002e-06, + "loss": 0.0347, + "step": 2902 + }, + { + "epoch": 3.4344868382135463, + "grad_norm": 1.467069387435913, + "learning_rate": 1.37125e-06, + "loss": 0.037, + "step": 2903 + }, + { + "epoch": 3.4356699201419696, + "grad_norm": 1.5232067108154297, + "learning_rate": 1.3700000000000002e-06, + "loss": 0.0308, + "step": 2904 + }, + { + "epoch": 3.4368530020703933, + "grad_norm": 1.4769593477249146, + "learning_rate": 1.36875e-06, + "loss": 0.0357, + "step": 2905 + }, + { + "epoch": 3.438036083998817, + "grad_norm": 1.5112065076828003, + "learning_rate": 1.3675000000000002e-06, + "loss": 0.0382, + "step": 2906 + }, + { + "epoch": 3.4392191659272404, + "grad_norm": 1.5942919254302979, + "learning_rate": 1.36625e-06, + "loss": 0.0371, + "step": 2907 + }, + { + "epoch": 3.440402247855664, + "grad_norm": 1.2236899137496948, + "learning_rate": 1.3650000000000003e-06, + "loss": 0.0347, + "step": 2908 + }, + { + "epoch": 3.4415853297840875, + "grad_norm": 1.5195512771606445, + "learning_rate": 1.36375e-06, + "loss": 0.033, + "step": 2909 + }, + { + "epoch": 3.4427684117125112, + "grad_norm": 1.702221393585205, + "learning_rate": 1.3625000000000003e-06, + "loss": 0.0316, + "step": 2910 + }, + { + "epoch": 3.4439514936409346, + "grad_norm": 1.6087162494659424, + "learning_rate": 1.36125e-06, + "loss": 0.0363, + "step": 2911 + }, + { + "epoch": 3.4451345755693583, + "grad_norm": 1.6247931718826294, + "learning_rate": 1.3600000000000001e-06, + "loss": 0.0389, + "step": 2912 + }, + { + "epoch": 3.4463176574977816, + "grad_norm": 1.429543137550354, + "learning_rate": 1.3587500000000001e-06, + "loss": 0.033, + "step": 2913 + }, + { + "epoch": 3.4475007394262054, + "grad_norm": 1.4637187719345093, + "learning_rate": 1.3575000000000001e-06, + "loss": 0.0321, + "step": 2914 + }, + { + "epoch": 3.4486838213546287, + "grad_norm": 1.5544235706329346, + "learning_rate": 1.3562500000000001e-06, + "loss": 0.0333, + "step": 2915 + }, + { + "epoch": 3.4498669032830525, + "grad_norm": 1.6133368015289307, + "learning_rate": 1.3550000000000002e-06, + "loss": 0.0312, + "step": 2916 + }, + { + "epoch": 3.4510499852114758, + "grad_norm": 1.2758756875991821, + "learning_rate": 1.3537500000000002e-06, + "loss": 0.0314, + "step": 2917 + }, + { + "epoch": 3.4522330671398995, + "grad_norm": 1.790597677230835, + "learning_rate": 1.3525000000000002e-06, + "loss": 0.0338, + "step": 2918 + }, + { + "epoch": 3.453416149068323, + "grad_norm": 1.3137996196746826, + "learning_rate": 1.35125e-06, + "loss": 0.0311, + "step": 2919 + }, + { + "epoch": 3.4545992309967466, + "grad_norm": 1.7594927549362183, + "learning_rate": 1.3500000000000002e-06, + "loss": 0.0369, + "step": 2920 + }, + { + "epoch": 3.45578231292517, + "grad_norm": 1.7023392915725708, + "learning_rate": 1.34875e-06, + "loss": 0.0376, + "step": 2921 + }, + { + "epoch": 3.4569653948535937, + "grad_norm": 1.5945905447006226, + "learning_rate": 1.3475000000000002e-06, + "loss": 0.0336, + "step": 2922 + }, + { + "epoch": 3.458148476782017, + "grad_norm": 1.7500190734863281, + "learning_rate": 1.34625e-06, + "loss": 0.0347, + "step": 2923 + }, + { + "epoch": 3.4593315587104407, + "grad_norm": 1.6093848943710327, + "learning_rate": 1.3450000000000003e-06, + "loss": 0.0296, + "step": 2924 + }, + { + "epoch": 3.460514640638864, + "grad_norm": 1.2678077220916748, + "learning_rate": 1.34375e-06, + "loss": 0.0304, + "step": 2925 + }, + { + "epoch": 3.461697722567288, + "grad_norm": 1.5084872245788574, + "learning_rate": 1.3425000000000003e-06, + "loss": 0.0392, + "step": 2926 + }, + { + "epoch": 3.462880804495711, + "grad_norm": 1.5005918741226196, + "learning_rate": 1.34125e-06, + "loss": 0.0373, + "step": 2927 + }, + { + "epoch": 3.464063886424135, + "grad_norm": 1.790027141571045, + "learning_rate": 1.34e-06, + "loss": 0.0329, + "step": 2928 + }, + { + "epoch": 3.4652469683525586, + "grad_norm": 1.5610442161560059, + "learning_rate": 1.33875e-06, + "loss": 0.0317, + "step": 2929 + }, + { + "epoch": 3.466430050280982, + "grad_norm": 1.4021284580230713, + "learning_rate": 1.3375000000000001e-06, + "loss": 0.0335, + "step": 2930 + }, + { + "epoch": 3.4676131322094053, + "grad_norm": 1.1816884279251099, + "learning_rate": 1.3362500000000001e-06, + "loss": 0.0357, + "step": 2931 + }, + { + "epoch": 3.468796214137829, + "grad_norm": 1.4016615152359009, + "learning_rate": 1.3350000000000001e-06, + "loss": 0.0356, + "step": 2932 + }, + { + "epoch": 3.4699792960662528, + "grad_norm": 1.532656192779541, + "learning_rate": 1.3337500000000001e-06, + "loss": 0.0334, + "step": 2933 + }, + { + "epoch": 3.471162377994676, + "grad_norm": 1.6098297834396362, + "learning_rate": 1.3325000000000002e-06, + "loss": 0.0318, + "step": 2934 + }, + { + "epoch": 3.4723454599231, + "grad_norm": 1.3510836362838745, + "learning_rate": 1.33125e-06, + "loss": 0.032, + "step": 2935 + }, + { + "epoch": 3.473528541851523, + "grad_norm": 1.329464316368103, + "learning_rate": 1.3300000000000002e-06, + "loss": 0.0355, + "step": 2936 + }, + { + "epoch": 3.474711623779947, + "grad_norm": 1.6773408651351929, + "learning_rate": 1.32875e-06, + "loss": 0.0329, + "step": 2937 + }, + { + "epoch": 3.4758947057083702, + "grad_norm": 1.59965181350708, + "learning_rate": 1.3275000000000002e-06, + "loss": 0.0367, + "step": 2938 + }, + { + "epoch": 3.477077787636794, + "grad_norm": 1.5206180810928345, + "learning_rate": 1.32625e-06, + "loss": 0.033, + "step": 2939 + }, + { + "epoch": 3.4782608695652173, + "grad_norm": 1.274746298789978, + "learning_rate": 1.3250000000000002e-06, + "loss": 0.0343, + "step": 2940 + }, + { + "epoch": 3.479443951493641, + "grad_norm": 1.4046766757965088, + "learning_rate": 1.32375e-06, + "loss": 0.0327, + "step": 2941 + }, + { + "epoch": 3.4806270334220644, + "grad_norm": 1.5636425018310547, + "learning_rate": 1.3225000000000003e-06, + "loss": 0.0363, + "step": 2942 + }, + { + "epoch": 3.481810115350488, + "grad_norm": 1.331530213356018, + "learning_rate": 1.32125e-06, + "loss": 0.0313, + "step": 2943 + }, + { + "epoch": 3.4829931972789114, + "grad_norm": 1.4865994453430176, + "learning_rate": 1.32e-06, + "loss": 0.0327, + "step": 2944 + }, + { + "epoch": 3.484176279207335, + "grad_norm": 1.3102092742919922, + "learning_rate": 1.31875e-06, + "loss": 0.0362, + "step": 2945 + }, + { + "epoch": 3.4853593611357585, + "grad_norm": 1.4523966312408447, + "learning_rate": 1.3175e-06, + "loss": 0.0314, + "step": 2946 + }, + { + "epoch": 3.4865424430641823, + "grad_norm": 1.2753907442092896, + "learning_rate": 1.31625e-06, + "loss": 0.0363, + "step": 2947 + }, + { + "epoch": 3.4877255249926056, + "grad_norm": 1.2386422157287598, + "learning_rate": 1.3150000000000001e-06, + "loss": 0.0335, + "step": 2948 + }, + { + "epoch": 3.4889086069210293, + "grad_norm": 1.6707817316055298, + "learning_rate": 1.3137500000000001e-06, + "loss": 0.0326, + "step": 2949 + }, + { + "epoch": 3.4900916888494526, + "grad_norm": 1.4847416877746582, + "learning_rate": 1.3125000000000001e-06, + "loss": 0.0373, + "step": 2950 + }, + { + "epoch": 3.4912747707778764, + "grad_norm": 1.5713376998901367, + "learning_rate": 1.31125e-06, + "loss": 0.0346, + "step": 2951 + }, + { + "epoch": 3.4924578527063, + "grad_norm": 1.1849539279937744, + "learning_rate": 1.3100000000000002e-06, + "loss": 0.034, + "step": 2952 + }, + { + "epoch": 3.4936409346347235, + "grad_norm": 1.5053311586380005, + "learning_rate": 1.30875e-06, + "loss": 0.0339, + "step": 2953 + }, + { + "epoch": 3.494824016563147, + "grad_norm": 1.3445355892181396, + "learning_rate": 1.3075000000000002e-06, + "loss": 0.0305, + "step": 2954 + }, + { + "epoch": 3.4960070984915705, + "grad_norm": 1.4720388650894165, + "learning_rate": 1.30625e-06, + "loss": 0.0341, + "step": 2955 + }, + { + "epoch": 3.4971901804199943, + "grad_norm": 1.4612491130828857, + "learning_rate": 1.3050000000000002e-06, + "loss": 0.0323, + "step": 2956 + }, + { + "epoch": 3.4983732623484176, + "grad_norm": 1.807178020477295, + "learning_rate": 1.30375e-06, + "loss": 0.0324, + "step": 2957 + }, + { + "epoch": 3.499556344276841, + "grad_norm": 1.3745554685592651, + "learning_rate": 1.3025000000000002e-06, + "loss": 0.034, + "step": 2958 + }, + { + "epoch": 3.5007394262052647, + "grad_norm": 1.310582160949707, + "learning_rate": 1.30125e-06, + "loss": 0.0342, + "step": 2959 + }, + { + "epoch": 3.5019225081336884, + "grad_norm": 1.4282643795013428, + "learning_rate": 1.3e-06, + "loss": 0.032, + "step": 2960 + }, + { + "epoch": 3.5031055900621118, + "grad_norm": 1.4388601779937744, + "learning_rate": 1.29875e-06, + "loss": 0.031, + "step": 2961 + }, + { + "epoch": 3.5042886719905355, + "grad_norm": 1.647520661354065, + "learning_rate": 1.2975e-06, + "loss": 0.0329, + "step": 2962 + }, + { + "epoch": 3.505471753918959, + "grad_norm": 1.653293490409851, + "learning_rate": 1.29625e-06, + "loss": 0.0331, + "step": 2963 + }, + { + "epoch": 3.5066548358473826, + "grad_norm": 1.5619944334030151, + "learning_rate": 1.295e-06, + "loss": 0.0393, + "step": 2964 + }, + { + "epoch": 3.507837917775806, + "grad_norm": 1.6929702758789062, + "learning_rate": 1.2937500000000001e-06, + "loss": 0.0342, + "step": 2965 + }, + { + "epoch": 3.5090209997042296, + "grad_norm": 1.3801772594451904, + "learning_rate": 1.2925000000000001e-06, + "loss": 0.0315, + "step": 2966 + }, + { + "epoch": 3.510204081632653, + "grad_norm": 1.3948968648910522, + "learning_rate": 1.29125e-06, + "loss": 0.0337, + "step": 2967 + }, + { + "epoch": 3.5113871635610767, + "grad_norm": 1.8774956464767456, + "learning_rate": 1.2900000000000001e-06, + "loss": 0.0354, + "step": 2968 + }, + { + "epoch": 3.5125702454895, + "grad_norm": 1.3476656675338745, + "learning_rate": 1.28875e-06, + "loss": 0.0328, + "step": 2969 + }, + { + "epoch": 3.513753327417924, + "grad_norm": 1.5439212322235107, + "learning_rate": 1.2875000000000002e-06, + "loss": 0.0357, + "step": 2970 + }, + { + "epoch": 3.514936409346347, + "grad_norm": 1.583813190460205, + "learning_rate": 1.28625e-06, + "loss": 0.0345, + "step": 2971 + }, + { + "epoch": 3.516119491274771, + "grad_norm": 1.7169080972671509, + "learning_rate": 1.2850000000000002e-06, + "loss": 0.0421, + "step": 2972 + }, + { + "epoch": 3.517302573203194, + "grad_norm": 1.4342660903930664, + "learning_rate": 1.28375e-06, + "loss": 0.032, + "step": 2973 + }, + { + "epoch": 3.518485655131618, + "grad_norm": 1.5545165538787842, + "learning_rate": 1.2825000000000002e-06, + "loss": 0.0367, + "step": 2974 + }, + { + "epoch": 3.5196687370600412, + "grad_norm": 2.1816046237945557, + "learning_rate": 1.28125e-06, + "loss": 0.0377, + "step": 2975 + }, + { + "epoch": 3.520851818988465, + "grad_norm": 1.3315095901489258, + "learning_rate": 1.28e-06, + "loss": 0.0335, + "step": 2976 + }, + { + "epoch": 3.5220349009168883, + "grad_norm": 1.539430856704712, + "learning_rate": 1.27875e-06, + "loss": 0.0383, + "step": 2977 + }, + { + "epoch": 3.523217982845312, + "grad_norm": 1.3595999479293823, + "learning_rate": 1.2775e-06, + "loss": 0.0338, + "step": 2978 + }, + { + "epoch": 3.524401064773736, + "grad_norm": 1.536086082458496, + "learning_rate": 1.27625e-06, + "loss": 0.0301, + "step": 2979 + }, + { + "epoch": 3.525584146702159, + "grad_norm": 1.3862330913543701, + "learning_rate": 1.275e-06, + "loss": 0.0379, + "step": 2980 + }, + { + "epoch": 3.5267672286305825, + "grad_norm": 1.6150269508361816, + "learning_rate": 1.27375e-06, + "loss": 0.0318, + "step": 2981 + }, + { + "epoch": 3.527950310559006, + "grad_norm": 1.6233856678009033, + "learning_rate": 1.2725e-06, + "loss": 0.0367, + "step": 2982 + }, + { + "epoch": 3.52913339248743, + "grad_norm": 1.5018876791000366, + "learning_rate": 1.27125e-06, + "loss": 0.0355, + "step": 2983 + }, + { + "epoch": 3.5303164744158533, + "grad_norm": 1.5526676177978516, + "learning_rate": 1.2700000000000001e-06, + "loss": 0.0361, + "step": 2984 + }, + { + "epoch": 3.5314995563442766, + "grad_norm": 1.6252474784851074, + "learning_rate": 1.26875e-06, + "loss": 0.0349, + "step": 2985 + }, + { + "epoch": 3.5326826382727003, + "grad_norm": 1.6919753551483154, + "learning_rate": 1.2675000000000001e-06, + "loss": 0.0335, + "step": 2986 + }, + { + "epoch": 3.533865720201124, + "grad_norm": 1.462986707687378, + "learning_rate": 1.26625e-06, + "loss": 0.0359, + "step": 2987 + }, + { + "epoch": 3.5350488021295474, + "grad_norm": 1.9608771800994873, + "learning_rate": 1.2650000000000002e-06, + "loss": 0.0423, + "step": 2988 + }, + { + "epoch": 3.536231884057971, + "grad_norm": 1.4618810415267944, + "learning_rate": 1.26375e-06, + "loss": 0.0363, + "step": 2989 + }, + { + "epoch": 3.5374149659863945, + "grad_norm": 1.3727037906646729, + "learning_rate": 1.2625000000000002e-06, + "loss": 0.0324, + "step": 2990 + }, + { + "epoch": 3.5385980479148182, + "grad_norm": 1.5059683322906494, + "learning_rate": 1.26125e-06, + "loss": 0.0348, + "step": 2991 + }, + { + "epoch": 3.5397811298432416, + "grad_norm": 1.938733458518982, + "learning_rate": 1.26e-06, + "loss": 0.0367, + "step": 2992 + }, + { + "epoch": 3.5409642117716653, + "grad_norm": 1.4872781038284302, + "learning_rate": 1.25875e-06, + "loss": 0.0338, + "step": 2993 + }, + { + "epoch": 3.5421472937000886, + "grad_norm": 1.3767532110214233, + "learning_rate": 1.2575e-06, + "loss": 0.0351, + "step": 2994 + }, + { + "epoch": 3.5433303756285124, + "grad_norm": 1.4074276685714722, + "learning_rate": 1.25625e-06, + "loss": 0.0335, + "step": 2995 + }, + { + "epoch": 3.5445134575569357, + "grad_norm": 1.6163203716278076, + "learning_rate": 1.255e-06, + "loss": 0.0348, + "step": 2996 + }, + { + "epoch": 3.5456965394853595, + "grad_norm": 1.766810417175293, + "learning_rate": 1.25375e-06, + "loss": 0.03, + "step": 2997 + }, + { + "epoch": 3.5468796214137828, + "grad_norm": 1.3383371829986572, + "learning_rate": 1.2525e-06, + "loss": 0.0335, + "step": 2998 + }, + { + "epoch": 3.5480627033422065, + "grad_norm": 1.4256573915481567, + "learning_rate": 1.2512499999999999e-06, + "loss": 0.0315, + "step": 2999 + }, + { + "epoch": 3.54924578527063, + "grad_norm": 1.556921124458313, + "learning_rate": 1.25e-06, + "loss": 0.0333, + "step": 3000 + }, + { + "epoch": 3.5504288671990536, + "grad_norm": 1.2363673448562622, + "learning_rate": 1.2487500000000001e-06, + "loss": 0.0321, + "step": 3001 + }, + { + "epoch": 3.5516119491274774, + "grad_norm": 1.5004466772079468, + "learning_rate": 1.2475000000000001e-06, + "loss": 0.0323, + "step": 3002 + }, + { + "epoch": 3.5527950310559007, + "grad_norm": 2.077425241470337, + "learning_rate": 1.2462500000000001e-06, + "loss": 0.0348, + "step": 3003 + }, + { + "epoch": 3.553978112984324, + "grad_norm": 1.5644367933273315, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.0334, + "step": 3004 + }, + { + "epoch": 3.5551611949127477, + "grad_norm": 1.5223509073257446, + "learning_rate": 1.2437500000000002e-06, + "loss": 0.0369, + "step": 3005 + }, + { + "epoch": 3.5563442768411715, + "grad_norm": 1.281394124031067, + "learning_rate": 1.2425000000000002e-06, + "loss": 0.0299, + "step": 3006 + }, + { + "epoch": 3.557527358769595, + "grad_norm": 1.3446863889694214, + "learning_rate": 1.2412500000000002e-06, + "loss": 0.0361, + "step": 3007 + }, + { + "epoch": 3.558710440698018, + "grad_norm": 1.415883183479309, + "learning_rate": 1.2400000000000002e-06, + "loss": 0.0341, + "step": 3008 + }, + { + "epoch": 3.559893522626442, + "grad_norm": 1.3649036884307861, + "learning_rate": 1.23875e-06, + "loss": 0.0303, + "step": 3009 + }, + { + "epoch": 3.5610766045548656, + "grad_norm": 1.6214039325714111, + "learning_rate": 1.2375e-06, + "loss": 0.0361, + "step": 3010 + }, + { + "epoch": 3.562259686483289, + "grad_norm": 1.774811863899231, + "learning_rate": 1.23625e-06, + "loss": 0.0364, + "step": 3011 + }, + { + "epoch": 3.5634427684117123, + "grad_norm": 1.4755815267562866, + "learning_rate": 1.235e-06, + "loss": 0.034, + "step": 3012 + }, + { + "epoch": 3.564625850340136, + "grad_norm": 1.462934970855713, + "learning_rate": 1.23375e-06, + "loss": 0.0308, + "step": 3013 + }, + { + "epoch": 3.5658089322685598, + "grad_norm": 1.415045142173767, + "learning_rate": 1.2325e-06, + "loss": 0.0347, + "step": 3014 + }, + { + "epoch": 3.566992014196983, + "grad_norm": 1.467208743095398, + "learning_rate": 1.23125e-06, + "loss": 0.0328, + "step": 3015 + }, + { + "epoch": 3.568175096125407, + "grad_norm": 1.4771373271942139, + "learning_rate": 1.23e-06, + "loss": 0.0332, + "step": 3016 + }, + { + "epoch": 3.56935817805383, + "grad_norm": 1.3425546884536743, + "learning_rate": 1.22875e-06, + "loss": 0.0316, + "step": 3017 + }, + { + "epoch": 3.570541259982254, + "grad_norm": 1.4445233345031738, + "learning_rate": 1.2275000000000001e-06, + "loss": 0.0345, + "step": 3018 + }, + { + "epoch": 3.5717243419106772, + "grad_norm": 1.5097103118896484, + "learning_rate": 1.2262500000000001e-06, + "loss": 0.034, + "step": 3019 + }, + { + "epoch": 3.572907423839101, + "grad_norm": 1.2671518325805664, + "learning_rate": 1.2250000000000001e-06, + "loss": 0.0301, + "step": 3020 + }, + { + "epoch": 3.5740905057675243, + "grad_norm": 1.492395281791687, + "learning_rate": 1.2237500000000001e-06, + "loss": 0.0365, + "step": 3021 + }, + { + "epoch": 3.575273587695948, + "grad_norm": 1.3370702266693115, + "learning_rate": 1.2225000000000002e-06, + "loss": 0.0335, + "step": 3022 + }, + { + "epoch": 3.5764566696243714, + "grad_norm": 1.3999507427215576, + "learning_rate": 1.2212500000000002e-06, + "loss": 0.0311, + "step": 3023 + }, + { + "epoch": 3.577639751552795, + "grad_norm": 1.4448567628860474, + "learning_rate": 1.2200000000000002e-06, + "loss": 0.0288, + "step": 3024 + }, + { + "epoch": 3.5788228334812184, + "grad_norm": 1.7021918296813965, + "learning_rate": 1.21875e-06, + "loss": 0.0383, + "step": 3025 + }, + { + "epoch": 3.580005915409642, + "grad_norm": 1.4945735931396484, + "learning_rate": 1.2175e-06, + "loss": 0.0344, + "step": 3026 + }, + { + "epoch": 3.5811889973380655, + "grad_norm": 1.5659414529800415, + "learning_rate": 1.21625e-06, + "loss": 0.0344, + "step": 3027 + }, + { + "epoch": 3.5823720792664893, + "grad_norm": 1.2697035074234009, + "learning_rate": 1.215e-06, + "loss": 0.032, + "step": 3028 + }, + { + "epoch": 3.583555161194913, + "grad_norm": 1.2943918704986572, + "learning_rate": 1.21375e-06, + "loss": 0.0364, + "step": 3029 + }, + { + "epoch": 3.5847382431233363, + "grad_norm": 1.451864242553711, + "learning_rate": 1.2125e-06, + "loss": 0.0342, + "step": 3030 + }, + { + "epoch": 3.5859213250517596, + "grad_norm": 1.5345979928970337, + "learning_rate": 1.21125e-06, + "loss": 0.0353, + "step": 3031 + }, + { + "epoch": 3.5871044069801834, + "grad_norm": 1.5697550773620605, + "learning_rate": 1.21e-06, + "loss": 0.0358, + "step": 3032 + }, + { + "epoch": 3.588287488908607, + "grad_norm": 1.471081018447876, + "learning_rate": 1.20875e-06, + "loss": 0.0331, + "step": 3033 + }, + { + "epoch": 3.5894705708370305, + "grad_norm": 1.5471588373184204, + "learning_rate": 1.2075e-06, + "loss": 0.03, + "step": 3034 + }, + { + "epoch": 3.590653652765454, + "grad_norm": 1.5841257572174072, + "learning_rate": 1.20625e-06, + "loss": 0.0371, + "step": 3035 + }, + { + "epoch": 3.5918367346938775, + "grad_norm": 1.0812010765075684, + "learning_rate": 1.2050000000000001e-06, + "loss": 0.034, + "step": 3036 + }, + { + "epoch": 3.5930198166223013, + "grad_norm": 1.701359510421753, + "learning_rate": 1.2037500000000001e-06, + "loss": 0.0359, + "step": 3037 + }, + { + "epoch": 3.5942028985507246, + "grad_norm": 1.7841334342956543, + "learning_rate": 1.2025000000000001e-06, + "loss": 0.0367, + "step": 3038 + }, + { + "epoch": 3.595385980479148, + "grad_norm": 1.8482797145843506, + "learning_rate": 1.2012500000000001e-06, + "loss": 0.0378, + "step": 3039 + }, + { + "epoch": 3.5965690624075717, + "grad_norm": 2.011932611465454, + "learning_rate": 1.2000000000000002e-06, + "loss": 0.0352, + "step": 3040 + }, + { + "epoch": 3.5977521443359954, + "grad_norm": 1.660944938659668, + "learning_rate": 1.19875e-06, + "loss": 0.0355, + "step": 3041 + }, + { + "epoch": 3.5989352262644188, + "grad_norm": 1.4434303045272827, + "learning_rate": 1.1975e-06, + "loss": 0.0356, + "step": 3042 + }, + { + "epoch": 3.6001183081928425, + "grad_norm": 1.4417921304702759, + "learning_rate": 1.19625e-06, + "loss": 0.031, + "step": 3043 + }, + { + "epoch": 3.601301390121266, + "grad_norm": 1.3762948513031006, + "learning_rate": 1.195e-06, + "loss": 0.0303, + "step": 3044 + }, + { + "epoch": 3.6024844720496896, + "grad_norm": 1.4027305841445923, + "learning_rate": 1.19375e-06, + "loss": 0.035, + "step": 3045 + }, + { + "epoch": 3.603667553978113, + "grad_norm": 1.5416052341461182, + "learning_rate": 1.1925e-06, + "loss": 0.031, + "step": 3046 + }, + { + "epoch": 3.6048506359065366, + "grad_norm": 1.273895025253296, + "learning_rate": 1.19125e-06, + "loss": 0.03, + "step": 3047 + }, + { + "epoch": 3.60603371783496, + "grad_norm": 1.2502799034118652, + "learning_rate": 1.19e-06, + "loss": 0.0293, + "step": 3048 + }, + { + "epoch": 3.6072167997633837, + "grad_norm": 1.8107908964157104, + "learning_rate": 1.18875e-06, + "loss": 0.0388, + "step": 3049 + }, + { + "epoch": 3.608399881691807, + "grad_norm": 1.5127145051956177, + "learning_rate": 1.1875e-06, + "loss": 0.0317, + "step": 3050 + }, + { + "epoch": 3.609582963620231, + "grad_norm": 1.6318848133087158, + "learning_rate": 1.18625e-06, + "loss": 0.0326, + "step": 3051 + }, + { + "epoch": 3.610766045548654, + "grad_norm": 1.413529396057129, + "learning_rate": 1.185e-06, + "loss": 0.0343, + "step": 3052 + }, + { + "epoch": 3.611949127477078, + "grad_norm": 1.4176348447799683, + "learning_rate": 1.18375e-06, + "loss": 0.0313, + "step": 3053 + }, + { + "epoch": 3.613132209405501, + "grad_norm": 1.6595139503479004, + "learning_rate": 1.1825000000000001e-06, + "loss": 0.0361, + "step": 3054 + }, + { + "epoch": 3.614315291333925, + "grad_norm": 1.5421091318130493, + "learning_rate": 1.1812500000000001e-06, + "loss": 0.0323, + "step": 3055 + }, + { + "epoch": 3.6154983732623487, + "grad_norm": 1.4150395393371582, + "learning_rate": 1.1800000000000001e-06, + "loss": 0.0311, + "step": 3056 + }, + { + "epoch": 3.616681455190772, + "grad_norm": 1.323448896408081, + "learning_rate": 1.17875e-06, + "loss": 0.0299, + "step": 3057 + }, + { + "epoch": 3.6178645371191953, + "grad_norm": 2.160806179046631, + "learning_rate": 1.1775e-06, + "loss": 0.0339, + "step": 3058 + }, + { + "epoch": 3.619047619047619, + "grad_norm": 1.5186008214950562, + "learning_rate": 1.17625e-06, + "loss": 0.0355, + "step": 3059 + }, + { + "epoch": 3.620230700976043, + "grad_norm": 1.4188816547393799, + "learning_rate": 1.175e-06, + "loss": 0.0315, + "step": 3060 + }, + { + "epoch": 3.621413782904466, + "grad_norm": 1.6154344081878662, + "learning_rate": 1.17375e-06, + "loss": 0.0368, + "step": 3061 + }, + { + "epoch": 3.6225968648328895, + "grad_norm": 1.3013142347335815, + "learning_rate": 1.1725e-06, + "loss": 0.0369, + "step": 3062 + }, + { + "epoch": 3.623779946761313, + "grad_norm": 1.3513195514678955, + "learning_rate": 1.1712500000000002e-06, + "loss": 0.0311, + "step": 3063 + }, + { + "epoch": 3.624963028689737, + "grad_norm": 1.5695844888687134, + "learning_rate": 1.1700000000000002e-06, + "loss": 0.0329, + "step": 3064 + }, + { + "epoch": 3.6261461106181603, + "grad_norm": 1.4686617851257324, + "learning_rate": 1.1687500000000002e-06, + "loss": 0.033, + "step": 3065 + }, + { + "epoch": 3.6273291925465836, + "grad_norm": 1.4151514768600464, + "learning_rate": 1.1675000000000003e-06, + "loss": 0.034, + "step": 3066 + }, + { + "epoch": 3.6285122744750073, + "grad_norm": 1.3076938390731812, + "learning_rate": 1.16625e-06, + "loss": 0.0289, + "step": 3067 + }, + { + "epoch": 3.629695356403431, + "grad_norm": 1.3568390607833862, + "learning_rate": 1.165e-06, + "loss": 0.0306, + "step": 3068 + }, + { + "epoch": 3.6308784383318544, + "grad_norm": 1.518460988998413, + "learning_rate": 1.16375e-06, + "loss": 0.0379, + "step": 3069 + }, + { + "epoch": 3.632061520260278, + "grad_norm": 1.4639091491699219, + "learning_rate": 1.1625e-06, + "loss": 0.0356, + "step": 3070 + }, + { + "epoch": 3.6332446021887015, + "grad_norm": 1.5994365215301514, + "learning_rate": 1.16125e-06, + "loss": 0.0303, + "step": 3071 + }, + { + "epoch": 3.6344276841171252, + "grad_norm": 1.5035464763641357, + "learning_rate": 1.1600000000000001e-06, + "loss": 0.0295, + "step": 3072 + }, + { + "epoch": 3.6356107660455486, + "grad_norm": 1.3631147146224976, + "learning_rate": 1.1587500000000001e-06, + "loss": 0.0358, + "step": 3073 + }, + { + "epoch": 3.6367938479739723, + "grad_norm": 1.3572660684585571, + "learning_rate": 1.1575000000000001e-06, + "loss": 0.0295, + "step": 3074 + }, + { + "epoch": 3.6379769299023956, + "grad_norm": 1.3243157863616943, + "learning_rate": 1.1562500000000002e-06, + "loss": 0.031, + "step": 3075 + }, + { + "epoch": 3.6391600118308194, + "grad_norm": 1.3842321634292603, + "learning_rate": 1.1550000000000002e-06, + "loss": 0.0324, + "step": 3076 + }, + { + "epoch": 3.6403430937592427, + "grad_norm": 1.2982652187347412, + "learning_rate": 1.1537500000000002e-06, + "loss": 0.0329, + "step": 3077 + }, + { + "epoch": 3.6415261756876665, + "grad_norm": 1.5167328119277954, + "learning_rate": 1.1525000000000002e-06, + "loss": 0.0353, + "step": 3078 + }, + { + "epoch": 3.6427092576160898, + "grad_norm": 1.526580572128296, + "learning_rate": 1.1512500000000002e-06, + "loss": 0.0398, + "step": 3079 + }, + { + "epoch": 3.6438923395445135, + "grad_norm": 1.3314740657806396, + "learning_rate": 1.1500000000000002e-06, + "loss": 0.0333, + "step": 3080 + }, + { + "epoch": 3.645075421472937, + "grad_norm": 1.4741747379302979, + "learning_rate": 1.1487500000000002e-06, + "loss": 0.0342, + "step": 3081 + }, + { + "epoch": 3.6462585034013606, + "grad_norm": 1.5334856510162354, + "learning_rate": 1.1475000000000002e-06, + "loss": 0.0368, + "step": 3082 + }, + { + "epoch": 3.6474415853297844, + "grad_norm": 1.3156050443649292, + "learning_rate": 1.14625e-06, + "loss": 0.0298, + "step": 3083 + }, + { + "epoch": 3.6486246672582077, + "grad_norm": 1.4871944189071655, + "learning_rate": 1.145e-06, + "loss": 0.0355, + "step": 3084 + }, + { + "epoch": 3.649807749186631, + "grad_norm": 1.340513825416565, + "learning_rate": 1.14375e-06, + "loss": 0.0363, + "step": 3085 + }, + { + "epoch": 3.6509908311150547, + "grad_norm": 1.885640263557434, + "learning_rate": 1.1425e-06, + "loss": 0.0329, + "step": 3086 + }, + { + "epoch": 3.6521739130434785, + "grad_norm": 1.306666612625122, + "learning_rate": 1.14125e-06, + "loss": 0.0319, + "step": 3087 + }, + { + "epoch": 3.653356994971902, + "grad_norm": 2.2113070487976074, + "learning_rate": 1.14e-06, + "loss": 0.0296, + "step": 3088 + }, + { + "epoch": 3.654540076900325, + "grad_norm": 1.6540635824203491, + "learning_rate": 1.1387500000000001e-06, + "loss": 0.0336, + "step": 3089 + }, + { + "epoch": 3.655723158828749, + "grad_norm": 1.3541659116744995, + "learning_rate": 1.1375000000000001e-06, + "loss": 0.0365, + "step": 3090 + }, + { + "epoch": 3.6569062407571726, + "grad_norm": 1.5614445209503174, + "learning_rate": 1.1362500000000001e-06, + "loss": 0.0324, + "step": 3091 + }, + { + "epoch": 3.658089322685596, + "grad_norm": 1.4541819095611572, + "learning_rate": 1.1350000000000001e-06, + "loss": 0.033, + "step": 3092 + }, + { + "epoch": 3.6592724046140193, + "grad_norm": 1.298319697380066, + "learning_rate": 1.1337500000000002e-06, + "loss": 0.0334, + "step": 3093 + }, + { + "epoch": 3.660455486542443, + "grad_norm": 1.4124705791473389, + "learning_rate": 1.1325000000000002e-06, + "loss": 0.0341, + "step": 3094 + }, + { + "epoch": 3.6616385684708668, + "grad_norm": 1.2728744745254517, + "learning_rate": 1.1312500000000002e-06, + "loss": 0.0327, + "step": 3095 + }, + { + "epoch": 3.66282165039929, + "grad_norm": 1.388566493988037, + "learning_rate": 1.1300000000000002e-06, + "loss": 0.0324, + "step": 3096 + }, + { + "epoch": 3.664004732327714, + "grad_norm": 1.8345109224319458, + "learning_rate": 1.1287500000000002e-06, + "loss": 0.033, + "step": 3097 + }, + { + "epoch": 3.665187814256137, + "grad_norm": 2.345740556716919, + "learning_rate": 1.1275000000000002e-06, + "loss": 0.039, + "step": 3098 + }, + { + "epoch": 3.666370896184561, + "grad_norm": 1.6117441654205322, + "learning_rate": 1.12625e-06, + "loss": 0.0332, + "step": 3099 + }, + { + "epoch": 3.6675539781129842, + "grad_norm": 1.4714605808258057, + "learning_rate": 1.125e-06, + "loss": 0.0307, + "step": 3100 + }, + { + "epoch": 3.668737060041408, + "grad_norm": 1.5398658514022827, + "learning_rate": 1.12375e-06, + "loss": 0.0342, + "step": 3101 + }, + { + "epoch": 3.6699201419698313, + "grad_norm": 1.5247820615768433, + "learning_rate": 1.1225e-06, + "loss": 0.0329, + "step": 3102 + }, + { + "epoch": 3.671103223898255, + "grad_norm": 1.4264508485794067, + "learning_rate": 1.12125e-06, + "loss": 0.0315, + "step": 3103 + }, + { + "epoch": 3.6722863058266784, + "grad_norm": 1.7256925106048584, + "learning_rate": 1.12e-06, + "loss": 0.0332, + "step": 3104 + }, + { + "epoch": 3.673469387755102, + "grad_norm": 1.3365249633789062, + "learning_rate": 1.11875e-06, + "loss": 0.036, + "step": 3105 + }, + { + "epoch": 3.6746524696835254, + "grad_norm": 1.3004368543624878, + "learning_rate": 1.1175e-06, + "loss": 0.0318, + "step": 3106 + }, + { + "epoch": 3.675835551611949, + "grad_norm": 1.257738471031189, + "learning_rate": 1.1162500000000001e-06, + "loss": 0.0323, + "step": 3107 + }, + { + "epoch": 3.6770186335403725, + "grad_norm": 1.5912891626358032, + "learning_rate": 1.1150000000000001e-06, + "loss": 0.0303, + "step": 3108 + }, + { + "epoch": 3.6782017154687963, + "grad_norm": 1.531211495399475, + "learning_rate": 1.1137500000000001e-06, + "loss": 0.0366, + "step": 3109 + }, + { + "epoch": 3.67938479739722, + "grad_norm": 1.2999118566513062, + "learning_rate": 1.1125000000000001e-06, + "loss": 0.0328, + "step": 3110 + }, + { + "epoch": 3.6805678793256433, + "grad_norm": 1.7561109066009521, + "learning_rate": 1.1112500000000002e-06, + "loss": 0.0372, + "step": 3111 + }, + { + "epoch": 3.6817509612540666, + "grad_norm": 1.3183891773223877, + "learning_rate": 1.1100000000000002e-06, + "loss": 0.0311, + "step": 3112 + }, + { + "epoch": 3.6829340431824904, + "grad_norm": 1.2963168621063232, + "learning_rate": 1.1087500000000002e-06, + "loss": 0.0348, + "step": 3113 + }, + { + "epoch": 3.684117125110914, + "grad_norm": 1.5430119037628174, + "learning_rate": 1.1075000000000002e-06, + "loss": 0.03, + "step": 3114 + }, + { + "epoch": 3.6853002070393375, + "grad_norm": 1.5069408416748047, + "learning_rate": 1.10625e-06, + "loss": 0.0294, + "step": 3115 + }, + { + "epoch": 3.686483288967761, + "grad_norm": 1.373327612876892, + "learning_rate": 1.105e-06, + "loss": 0.032, + "step": 3116 + }, + { + "epoch": 3.6876663708961845, + "grad_norm": 1.4403420686721802, + "learning_rate": 1.10375e-06, + "loss": 0.0333, + "step": 3117 + }, + { + "epoch": 3.6888494528246083, + "grad_norm": 1.371446967124939, + "learning_rate": 1.1025e-06, + "loss": 0.0309, + "step": 3118 + }, + { + "epoch": 3.6900325347530316, + "grad_norm": 1.562498688697815, + "learning_rate": 1.10125e-06, + "loss": 0.0326, + "step": 3119 + }, + { + "epoch": 3.691215616681455, + "grad_norm": 1.4720321893692017, + "learning_rate": 1.1e-06, + "loss": 0.0305, + "step": 3120 + }, + { + "epoch": 3.6923986986098787, + "grad_norm": 1.4292125701904297, + "learning_rate": 1.09875e-06, + "loss": 0.0339, + "step": 3121 + }, + { + "epoch": 3.6935817805383024, + "grad_norm": 1.5458494424819946, + "learning_rate": 1.0975e-06, + "loss": 0.0377, + "step": 3122 + }, + { + "epoch": 3.6947648624667258, + "grad_norm": 1.4622817039489746, + "learning_rate": 1.09625e-06, + "loss": 0.0278, + "step": 3123 + }, + { + "epoch": 3.6959479443951495, + "grad_norm": 1.5361117124557495, + "learning_rate": 1.095e-06, + "loss": 0.0318, + "step": 3124 + }, + { + "epoch": 3.697131026323573, + "grad_norm": 1.5040806531906128, + "learning_rate": 1.0937500000000001e-06, + "loss": 0.0333, + "step": 3125 + }, + { + "epoch": 3.6983141082519966, + "grad_norm": 1.441792368888855, + "learning_rate": 1.0925000000000001e-06, + "loss": 0.0335, + "step": 3126 + }, + { + "epoch": 3.69949719018042, + "grad_norm": 1.6298925876617432, + "learning_rate": 1.0912500000000001e-06, + "loss": 0.0309, + "step": 3127 + }, + { + "epoch": 3.7006802721088436, + "grad_norm": 1.4906195402145386, + "learning_rate": 1.0900000000000002e-06, + "loss": 0.0356, + "step": 3128 + }, + { + "epoch": 3.701863354037267, + "grad_norm": 1.4159423112869263, + "learning_rate": 1.0887500000000002e-06, + "loss": 0.0364, + "step": 3129 + }, + { + "epoch": 3.7030464359656907, + "grad_norm": 1.4641478061676025, + "learning_rate": 1.0875000000000002e-06, + "loss": 0.0306, + "step": 3130 + }, + { + "epoch": 3.704229517894114, + "grad_norm": 1.105021595954895, + "learning_rate": 1.0862500000000002e-06, + "loss": 0.0267, + "step": 3131 + }, + { + "epoch": 3.705412599822538, + "grad_norm": 1.540403962135315, + "learning_rate": 1.085e-06, + "loss": 0.0353, + "step": 3132 + }, + { + "epoch": 3.706595681750961, + "grad_norm": 1.4513791799545288, + "learning_rate": 1.08375e-06, + "loss": 0.0295, + "step": 3133 + }, + { + "epoch": 3.707778763679385, + "grad_norm": 1.3991225957870483, + "learning_rate": 1.0825e-06, + "loss": 0.0296, + "step": 3134 + }, + { + "epoch": 3.708961845607808, + "grad_norm": 1.8403664827346802, + "learning_rate": 1.08125e-06, + "loss": 0.0319, + "step": 3135 + }, + { + "epoch": 3.710144927536232, + "grad_norm": 1.3372331857681274, + "learning_rate": 1.08e-06, + "loss": 0.0321, + "step": 3136 + }, + { + "epoch": 3.7113280094646557, + "grad_norm": 1.8407433032989502, + "learning_rate": 1.07875e-06, + "loss": 0.0335, + "step": 3137 + }, + { + "epoch": 3.712511091393079, + "grad_norm": 1.8109571933746338, + "learning_rate": 1.0775e-06, + "loss": 0.0339, + "step": 3138 + }, + { + "epoch": 3.7136941733215023, + "grad_norm": 1.6841219663619995, + "learning_rate": 1.07625e-06, + "loss": 0.0359, + "step": 3139 + }, + { + "epoch": 3.714877255249926, + "grad_norm": 1.4224114418029785, + "learning_rate": 1.075e-06, + "loss": 0.033, + "step": 3140 + }, + { + "epoch": 3.71606033717835, + "grad_norm": 1.4705008268356323, + "learning_rate": 1.07375e-06, + "loss": 0.034, + "step": 3141 + }, + { + "epoch": 3.717243419106773, + "grad_norm": 1.2502061128616333, + "learning_rate": 1.0725000000000001e-06, + "loss": 0.0336, + "step": 3142 + }, + { + "epoch": 3.7184265010351965, + "grad_norm": 1.4459385871887207, + "learning_rate": 1.0712500000000001e-06, + "loss": 0.0327, + "step": 3143 + }, + { + "epoch": 3.71960958296362, + "grad_norm": 1.6611433029174805, + "learning_rate": 1.0700000000000001e-06, + "loss": 0.0343, + "step": 3144 + }, + { + "epoch": 3.720792664892044, + "grad_norm": 1.4321211576461792, + "learning_rate": 1.0687500000000001e-06, + "loss": 0.0325, + "step": 3145 + }, + { + "epoch": 3.7219757468204673, + "grad_norm": 1.3198440074920654, + "learning_rate": 1.0675000000000002e-06, + "loss": 0.0286, + "step": 3146 + }, + { + "epoch": 3.7231588287488906, + "grad_norm": 1.5870091915130615, + "learning_rate": 1.0662500000000002e-06, + "loss": 0.0326, + "step": 3147 + }, + { + "epoch": 3.7243419106773143, + "grad_norm": 1.4230053424835205, + "learning_rate": 1.065e-06, + "loss": 0.0287, + "step": 3148 + }, + { + "epoch": 3.725524992605738, + "grad_norm": 1.663040280342102, + "learning_rate": 1.06375e-06, + "loss": 0.0339, + "step": 3149 + }, + { + "epoch": 3.7267080745341614, + "grad_norm": 1.4405732154846191, + "learning_rate": 1.0625e-06, + "loss": 0.0339, + "step": 3150 + }, + { + "epoch": 3.727891156462585, + "grad_norm": 1.4021989107131958, + "learning_rate": 1.06125e-06, + "loss": 0.0298, + "step": 3151 + }, + { + "epoch": 3.7290742383910085, + "grad_norm": 1.3265639543533325, + "learning_rate": 1.06e-06, + "loss": 0.0334, + "step": 3152 + }, + { + "epoch": 3.7302573203194322, + "grad_norm": 2.3076698780059814, + "learning_rate": 1.05875e-06, + "loss": 0.0356, + "step": 3153 + }, + { + "epoch": 3.7314404022478556, + "grad_norm": 1.401691198348999, + "learning_rate": 1.0575e-06, + "loss": 0.0291, + "step": 3154 + }, + { + "epoch": 3.7326234841762793, + "grad_norm": 1.3240450620651245, + "learning_rate": 1.05625e-06, + "loss": 0.0306, + "step": 3155 + }, + { + "epoch": 3.7338065661047026, + "grad_norm": 1.481643795967102, + "learning_rate": 1.055e-06, + "loss": 0.0348, + "step": 3156 + }, + { + "epoch": 3.7349896480331264, + "grad_norm": 1.2509660720825195, + "learning_rate": 1.05375e-06, + "loss": 0.0306, + "step": 3157 + }, + { + "epoch": 3.7361727299615497, + "grad_norm": 1.504357099533081, + "learning_rate": 1.0525e-06, + "loss": 0.0368, + "step": 3158 + }, + { + "epoch": 3.7373558118899735, + "grad_norm": 1.4235472679138184, + "learning_rate": 1.05125e-06, + "loss": 0.0322, + "step": 3159 + }, + { + "epoch": 3.7385388938183968, + "grad_norm": 1.3889226913452148, + "learning_rate": 1.0500000000000001e-06, + "loss": 0.0267, + "step": 3160 + }, + { + "epoch": 3.7397219757468205, + "grad_norm": 1.4888522624969482, + "learning_rate": 1.0487500000000001e-06, + "loss": 0.0326, + "step": 3161 + }, + { + "epoch": 3.740905057675244, + "grad_norm": 1.452012062072754, + "learning_rate": 1.0475000000000001e-06, + "loss": 0.0329, + "step": 3162 + }, + { + "epoch": 3.7420881396036676, + "grad_norm": 1.3943297863006592, + "learning_rate": 1.0462500000000001e-06, + "loss": 0.0354, + "step": 3163 + }, + { + "epoch": 3.7432712215320914, + "grad_norm": 1.699652910232544, + "learning_rate": 1.045e-06, + "loss": 0.0333, + "step": 3164 + }, + { + "epoch": 3.7444543034605147, + "grad_norm": 1.4278453588485718, + "learning_rate": 1.04375e-06, + "loss": 0.0344, + "step": 3165 + }, + { + "epoch": 3.745637385388938, + "grad_norm": 1.4744415283203125, + "learning_rate": 1.0425e-06, + "loss": 0.0334, + "step": 3166 + }, + { + "epoch": 3.7468204673173617, + "grad_norm": 1.9033408164978027, + "learning_rate": 1.04125e-06, + "loss": 0.0316, + "step": 3167 + }, + { + "epoch": 3.7480035492457855, + "grad_norm": 1.2885016202926636, + "learning_rate": 1.04e-06, + "loss": 0.0318, + "step": 3168 + }, + { + "epoch": 3.749186631174209, + "grad_norm": 2.6570677757263184, + "learning_rate": 1.03875e-06, + "loss": 0.0344, + "step": 3169 + }, + { + "epoch": 3.750369713102632, + "grad_norm": 1.5496710538864136, + "learning_rate": 1.0375e-06, + "loss": 0.0346, + "step": 3170 + }, + { + "epoch": 3.751552795031056, + "grad_norm": 1.2244137525558472, + "learning_rate": 1.03625e-06, + "loss": 0.0293, + "step": 3171 + }, + { + "epoch": 3.7527358769594796, + "grad_norm": 1.537782907485962, + "learning_rate": 1.035e-06, + "loss": 0.0333, + "step": 3172 + }, + { + "epoch": 3.753918958887903, + "grad_norm": 1.2989498376846313, + "learning_rate": 1.03375e-06, + "loss": 0.0334, + "step": 3173 + }, + { + "epoch": 3.7551020408163263, + "grad_norm": 1.3052787780761719, + "learning_rate": 1.0325e-06, + "loss": 0.033, + "step": 3174 + }, + { + "epoch": 3.75628512274475, + "grad_norm": 1.4153378009796143, + "learning_rate": 1.03125e-06, + "loss": 0.027, + "step": 3175 + }, + { + "epoch": 3.7574682046731738, + "grad_norm": 1.5986751317977905, + "learning_rate": 1.03e-06, + "loss": 0.0321, + "step": 3176 + }, + { + "epoch": 3.758651286601597, + "grad_norm": 1.1900631189346313, + "learning_rate": 1.02875e-06, + "loss": 0.0321, + "step": 3177 + }, + { + "epoch": 3.759834368530021, + "grad_norm": 1.6101644039154053, + "learning_rate": 1.0275000000000001e-06, + "loss": 0.0374, + "step": 3178 + }, + { + "epoch": 3.761017450458444, + "grad_norm": 1.3567906618118286, + "learning_rate": 1.0262500000000001e-06, + "loss": 0.0317, + "step": 3179 + }, + { + "epoch": 3.762200532386868, + "grad_norm": 1.6403671503067017, + "learning_rate": 1.025e-06, + "loss": 0.0308, + "step": 3180 + }, + { + "epoch": 3.7633836143152912, + "grad_norm": 1.4633855819702148, + "learning_rate": 1.02375e-06, + "loss": 0.0309, + "step": 3181 + }, + { + "epoch": 3.764566696243715, + "grad_norm": 1.488281011581421, + "learning_rate": 1.0225e-06, + "loss": 0.0303, + "step": 3182 + }, + { + "epoch": 3.7657497781721383, + "grad_norm": 1.6326606273651123, + "learning_rate": 1.02125e-06, + "loss": 0.0332, + "step": 3183 + }, + { + "epoch": 3.766932860100562, + "grad_norm": 1.6419352293014526, + "learning_rate": 1.02e-06, + "loss": 0.0302, + "step": 3184 + }, + { + "epoch": 3.7681159420289854, + "grad_norm": 1.358381986618042, + "learning_rate": 1.01875e-06, + "loss": 0.0362, + "step": 3185 + }, + { + "epoch": 3.769299023957409, + "grad_norm": 1.3346877098083496, + "learning_rate": 1.0175e-06, + "loss": 0.0329, + "step": 3186 + }, + { + "epoch": 3.7704821058858324, + "grad_norm": 1.6784603595733643, + "learning_rate": 1.01625e-06, + "loss": 0.0405, + "step": 3187 + }, + { + "epoch": 3.771665187814256, + "grad_norm": 1.8583189249038696, + "learning_rate": 1.0150000000000002e-06, + "loss": 0.0317, + "step": 3188 + }, + { + "epoch": 3.7728482697426795, + "grad_norm": 1.517810344696045, + "learning_rate": 1.0137500000000002e-06, + "loss": 0.0313, + "step": 3189 + }, + { + "epoch": 3.7740313516711033, + "grad_norm": 1.5661742687225342, + "learning_rate": 1.0125e-06, + "loss": 0.0334, + "step": 3190 + }, + { + "epoch": 3.775214433599527, + "grad_norm": 1.166453242301941, + "learning_rate": 1.01125e-06, + "loss": 0.036, + "step": 3191 + }, + { + "epoch": 3.7763975155279503, + "grad_norm": 1.5571891069412231, + "learning_rate": 1.01e-06, + "loss": 0.0364, + "step": 3192 + }, + { + "epoch": 3.7775805974563736, + "grad_norm": 1.3530018329620361, + "learning_rate": 1.00875e-06, + "loss": 0.0344, + "step": 3193 + }, + { + "epoch": 3.7787636793847974, + "grad_norm": 1.558789849281311, + "learning_rate": 1.0075e-06, + "loss": 0.0398, + "step": 3194 + }, + { + "epoch": 3.779946761313221, + "grad_norm": 1.419532060623169, + "learning_rate": 1.0062500000000001e-06, + "loss": 0.0353, + "step": 3195 + }, + { + "epoch": 3.7811298432416445, + "grad_norm": 1.333701252937317, + "learning_rate": 1.0050000000000001e-06, + "loss": 0.0311, + "step": 3196 + }, + { + "epoch": 3.782312925170068, + "grad_norm": 1.2630265951156616, + "learning_rate": 1.0037500000000001e-06, + "loss": 0.0328, + "step": 3197 + }, + { + "epoch": 3.7834960070984915, + "grad_norm": 1.6110697984695435, + "learning_rate": 1.0025000000000001e-06, + "loss": 0.0353, + "step": 3198 + }, + { + "epoch": 3.7846790890269153, + "grad_norm": 2.208242654800415, + "learning_rate": 1.0012500000000002e-06, + "loss": 0.0362, + "step": 3199 + }, + { + "epoch": 3.7858621709553386, + "grad_norm": 1.1792398691177368, + "learning_rate": 1.0000000000000002e-06, + "loss": 0.0317, + "step": 3200 + }, + { + "epoch": 3.787045252883762, + "grad_norm": 1.5717880725860596, + "learning_rate": 9.987500000000002e-07, + "loss": 0.033, + "step": 3201 + }, + { + "epoch": 3.7882283348121857, + "grad_norm": 1.2913737297058105, + "learning_rate": 9.975000000000002e-07, + "loss": 0.0313, + "step": 3202 + }, + { + "epoch": 3.7894114167406094, + "grad_norm": 1.3511567115783691, + "learning_rate": 9.962500000000002e-07, + "loss": 0.032, + "step": 3203 + }, + { + "epoch": 3.7905944986690328, + "grad_norm": 1.430700421333313, + "learning_rate": 9.950000000000002e-07, + "loss": 0.0362, + "step": 3204 + }, + { + "epoch": 3.7917775805974565, + "grad_norm": 1.5573371648788452, + "learning_rate": 9.937500000000002e-07, + "loss": 0.0346, + "step": 3205 + }, + { + "epoch": 3.79296066252588, + "grad_norm": 1.3145581483840942, + "learning_rate": 9.925e-07, + "loss": 0.0309, + "step": 3206 + }, + { + "epoch": 3.7941437444543036, + "grad_norm": 1.3016940355300903, + "learning_rate": 9.9125e-07, + "loss": 0.0346, + "step": 3207 + }, + { + "epoch": 3.795326826382727, + "grad_norm": 1.1285688877105713, + "learning_rate": 9.9e-07, + "loss": 0.0311, + "step": 3208 + }, + { + "epoch": 3.7965099083111506, + "grad_norm": 1.6230084896087646, + "learning_rate": 9.8875e-07, + "loss": 0.0324, + "step": 3209 + }, + { + "epoch": 3.797692990239574, + "grad_norm": 1.6754345893859863, + "learning_rate": 9.875e-07, + "loss": 0.0331, + "step": 3210 + }, + { + "epoch": 3.7988760721679977, + "grad_norm": 1.6067066192626953, + "learning_rate": 9.8625e-07, + "loss": 0.0328, + "step": 3211 + }, + { + "epoch": 3.800059154096421, + "grad_norm": 1.6142659187316895, + "learning_rate": 9.85e-07, + "loss": 0.0322, + "step": 3212 + }, + { + "epoch": 3.801242236024845, + "grad_norm": 1.5578984022140503, + "learning_rate": 9.837500000000001e-07, + "loss": 0.0371, + "step": 3213 + }, + { + "epoch": 3.802425317953268, + "grad_norm": 1.4418067932128906, + "learning_rate": 9.825000000000001e-07, + "loss": 0.0316, + "step": 3214 + }, + { + "epoch": 3.803608399881692, + "grad_norm": 1.4057230949401855, + "learning_rate": 9.812500000000001e-07, + "loss": 0.0315, + "step": 3215 + }, + { + "epoch": 3.804791481810115, + "grad_norm": 1.4127693176269531, + "learning_rate": 9.800000000000001e-07, + "loss": 0.0314, + "step": 3216 + }, + { + "epoch": 3.805974563738539, + "grad_norm": 1.3902525901794434, + "learning_rate": 9.787500000000002e-07, + "loss": 0.0328, + "step": 3217 + }, + { + "epoch": 3.8071576456669627, + "grad_norm": 1.4748363494873047, + "learning_rate": 9.775000000000002e-07, + "loss": 0.0271, + "step": 3218 + }, + { + "epoch": 3.808340727595386, + "grad_norm": 1.344033122062683, + "learning_rate": 9.762500000000002e-07, + "loss": 0.0269, + "step": 3219 + }, + { + "epoch": 3.8095238095238093, + "grad_norm": 1.570331335067749, + "learning_rate": 9.750000000000002e-07, + "loss": 0.0358, + "step": 3220 + }, + { + "epoch": 3.810706891452233, + "grad_norm": 1.4830741882324219, + "learning_rate": 9.737500000000002e-07, + "loss": 0.0384, + "step": 3221 + }, + { + "epoch": 3.811889973380657, + "grad_norm": 1.634732723236084, + "learning_rate": 9.725e-07, + "loss": 0.0331, + "step": 3222 + }, + { + "epoch": 3.81307305530908, + "grad_norm": 1.7948204278945923, + "learning_rate": 9.7125e-07, + "loss": 0.0338, + "step": 3223 + }, + { + "epoch": 3.8142561372375035, + "grad_norm": 1.5373040437698364, + "learning_rate": 9.7e-07, + "loss": 0.0311, + "step": 3224 + }, + { + "epoch": 3.815439219165927, + "grad_norm": 1.6260422468185425, + "learning_rate": 9.6875e-07, + "loss": 0.0362, + "step": 3225 + }, + { + "epoch": 3.816622301094351, + "grad_norm": 1.4102494716644287, + "learning_rate": 9.675e-07, + "loss": 0.0311, + "step": 3226 + }, + { + "epoch": 3.8178053830227743, + "grad_norm": 1.5827268362045288, + "learning_rate": 9.6625e-07, + "loss": 0.0312, + "step": 3227 + }, + { + "epoch": 3.818988464951198, + "grad_norm": 1.4003534317016602, + "learning_rate": 9.65e-07, + "loss": 0.0274, + "step": 3228 + }, + { + "epoch": 3.8201715468796213, + "grad_norm": 1.661650538444519, + "learning_rate": 9.6375e-07, + "loss": 0.0395, + "step": 3229 + }, + { + "epoch": 3.821354628808045, + "grad_norm": 1.214439868927002, + "learning_rate": 9.625e-07, + "loss": 0.0293, + "step": 3230 + }, + { + "epoch": 3.8225377107364684, + "grad_norm": 1.35824453830719, + "learning_rate": 9.612500000000001e-07, + "loss": 0.0271, + "step": 3231 + }, + { + "epoch": 3.823720792664892, + "grad_norm": 1.4084033966064453, + "learning_rate": 9.600000000000001e-07, + "loss": 0.03, + "step": 3232 + }, + { + "epoch": 3.8249038745933155, + "grad_norm": 1.346104383468628, + "learning_rate": 9.587500000000001e-07, + "loss": 0.033, + "step": 3233 + }, + { + "epoch": 3.8260869565217392, + "grad_norm": 1.3276923894882202, + "learning_rate": 9.575000000000001e-07, + "loss": 0.0291, + "step": 3234 + }, + { + "epoch": 3.8272700384501626, + "grad_norm": 1.35928213596344, + "learning_rate": 9.562500000000002e-07, + "loss": 0.0337, + "step": 3235 + }, + { + "epoch": 3.8284531203785863, + "grad_norm": 1.5496981143951416, + "learning_rate": 9.550000000000002e-07, + "loss": 0.0329, + "step": 3236 + }, + { + "epoch": 3.8296362023070096, + "grad_norm": 1.156783103942871, + "learning_rate": 9.537500000000002e-07, + "loss": 0.0293, + "step": 3237 + }, + { + "epoch": 3.8308192842354334, + "grad_norm": 1.6587222814559937, + "learning_rate": 9.525000000000001e-07, + "loss": 0.0312, + "step": 3238 + }, + { + "epoch": 3.8320023661638567, + "grad_norm": 1.3916010856628418, + "learning_rate": 9.512500000000001e-07, + "loss": 0.0335, + "step": 3239 + }, + { + "epoch": 3.8331854480922805, + "grad_norm": 1.244348406791687, + "learning_rate": 9.500000000000001e-07, + "loss": 0.028, + "step": 3240 + }, + { + "epoch": 3.8343685300207038, + "grad_norm": 1.5186567306518555, + "learning_rate": 9.487500000000001e-07, + "loss": 0.0322, + "step": 3241 + }, + { + "epoch": 3.8355516119491275, + "grad_norm": 1.4650102853775024, + "learning_rate": 9.475e-07, + "loss": 0.0289, + "step": 3242 + }, + { + "epoch": 3.836734693877551, + "grad_norm": 1.9524774551391602, + "learning_rate": 9.4625e-07, + "loss": 0.0334, + "step": 3243 + }, + { + "epoch": 3.8379177758059746, + "grad_norm": 1.3681143522262573, + "learning_rate": 9.450000000000001e-07, + "loss": 0.0366, + "step": 3244 + }, + { + "epoch": 3.8391008577343984, + "grad_norm": 1.5938427448272705, + "learning_rate": 9.437500000000001e-07, + "loss": 0.0332, + "step": 3245 + }, + { + "epoch": 3.8402839396628217, + "grad_norm": 1.5586495399475098, + "learning_rate": 9.425000000000001e-07, + "loss": 0.0366, + "step": 3246 + }, + { + "epoch": 3.841467021591245, + "grad_norm": 1.6967747211456299, + "learning_rate": 9.412500000000001e-07, + "loss": 0.0346, + "step": 3247 + }, + { + "epoch": 3.8426501035196687, + "grad_norm": 1.5958651304244995, + "learning_rate": 9.400000000000001e-07, + "loss": 0.036, + "step": 3248 + }, + { + "epoch": 3.8438331854480925, + "grad_norm": 1.6773791313171387, + "learning_rate": 9.387500000000001e-07, + "loss": 0.0322, + "step": 3249 + }, + { + "epoch": 3.845016267376516, + "grad_norm": 1.6645792722702026, + "learning_rate": 9.375000000000001e-07, + "loss": 0.0347, + "step": 3250 + }, + { + "epoch": 3.846199349304939, + "grad_norm": 1.4369484186172485, + "learning_rate": 9.3625e-07, + "loss": 0.0348, + "step": 3251 + }, + { + "epoch": 3.847382431233363, + "grad_norm": 1.5345337390899658, + "learning_rate": 9.35e-07, + "loss": 0.0328, + "step": 3252 + }, + { + "epoch": 3.8485655131617866, + "grad_norm": 1.382237195968628, + "learning_rate": 9.337500000000001e-07, + "loss": 0.0292, + "step": 3253 + }, + { + "epoch": 3.84974859509021, + "grad_norm": 1.4255424737930298, + "learning_rate": 9.325000000000001e-07, + "loss": 0.0358, + "step": 3254 + }, + { + "epoch": 3.8509316770186337, + "grad_norm": 1.3793935775756836, + "learning_rate": 9.312500000000001e-07, + "loss": 0.03, + "step": 3255 + }, + { + "epoch": 3.852114758947057, + "grad_norm": 1.8250515460968018, + "learning_rate": 9.300000000000001e-07, + "loss": 0.033, + "step": 3256 + }, + { + "epoch": 3.8532978408754808, + "grad_norm": 1.4007539749145508, + "learning_rate": 9.287500000000001e-07, + "loss": 0.0317, + "step": 3257 + }, + { + "epoch": 3.854480922803904, + "grad_norm": 1.4631458520889282, + "learning_rate": 9.275000000000001e-07, + "loss": 0.0336, + "step": 3258 + }, + { + "epoch": 3.855664004732328, + "grad_norm": 1.438245177268982, + "learning_rate": 9.2625e-07, + "loss": 0.0349, + "step": 3259 + }, + { + "epoch": 3.856847086660751, + "grad_norm": 1.9849612712860107, + "learning_rate": 9.25e-07, + "loss": 0.0316, + "step": 3260 + }, + { + "epoch": 3.858030168589175, + "grad_norm": 1.3967581987380981, + "learning_rate": 9.2375e-07, + "loss": 0.033, + "step": 3261 + }, + { + "epoch": 3.8592132505175982, + "grad_norm": 1.5747380256652832, + "learning_rate": 9.225000000000001e-07, + "loss": 0.0418, + "step": 3262 + }, + { + "epoch": 3.860396332446022, + "grad_norm": 1.4264438152313232, + "learning_rate": 9.212500000000001e-07, + "loss": 0.0367, + "step": 3263 + }, + { + "epoch": 3.8615794143744453, + "grad_norm": 2.016354560852051, + "learning_rate": 9.200000000000001e-07, + "loss": 0.0385, + "step": 3264 + }, + { + "epoch": 3.862762496302869, + "grad_norm": 1.1110016107559204, + "learning_rate": 9.187500000000001e-07, + "loss": 0.0301, + "step": 3265 + }, + { + "epoch": 3.8639455782312924, + "grad_norm": 1.610900640487671, + "learning_rate": 9.175000000000001e-07, + "loss": 0.0333, + "step": 3266 + }, + { + "epoch": 3.865128660159716, + "grad_norm": 1.4013073444366455, + "learning_rate": 9.1625e-07, + "loss": 0.0329, + "step": 3267 + }, + { + "epoch": 3.86631174208814, + "grad_norm": 1.5782734155654907, + "learning_rate": 9.15e-07, + "loss": 0.0319, + "step": 3268 + }, + { + "epoch": 3.867494824016563, + "grad_norm": 1.3669109344482422, + "learning_rate": 9.1375e-07, + "loss": 0.0366, + "step": 3269 + }, + { + "epoch": 3.8686779059449865, + "grad_norm": 1.6442350149154663, + "learning_rate": 9.125e-07, + "loss": 0.0312, + "step": 3270 + }, + { + "epoch": 3.8698609878734103, + "grad_norm": 1.2196000814437866, + "learning_rate": 9.112500000000001e-07, + "loss": 0.0405, + "step": 3271 + }, + { + "epoch": 3.871044069801834, + "grad_norm": 1.338904857635498, + "learning_rate": 9.100000000000001e-07, + "loss": 0.0326, + "step": 3272 + }, + { + "epoch": 3.8722271517302573, + "grad_norm": 1.4010568857192993, + "learning_rate": 9.087500000000001e-07, + "loss": 0.0296, + "step": 3273 + }, + { + "epoch": 3.8734102336586806, + "grad_norm": 1.350022315979004, + "learning_rate": 9.075000000000001e-07, + "loss": 0.0311, + "step": 3274 + }, + { + "epoch": 3.8745933155871044, + "grad_norm": 1.7890986204147339, + "learning_rate": 9.0625e-07, + "loss": 0.0341, + "step": 3275 + }, + { + "epoch": 3.875776397515528, + "grad_norm": 1.3625928163528442, + "learning_rate": 9.05e-07, + "loss": 0.028, + "step": 3276 + }, + { + "epoch": 3.8769594794439515, + "grad_norm": 1.8094254732131958, + "learning_rate": 9.0375e-07, + "loss": 0.0337, + "step": 3277 + }, + { + "epoch": 3.878142561372375, + "grad_norm": 1.27149498462677, + "learning_rate": 9.025e-07, + "loss": 0.0349, + "step": 3278 + }, + { + "epoch": 3.8793256433007985, + "grad_norm": 1.318089246749878, + "learning_rate": 9.012500000000001e-07, + "loss": 0.0312, + "step": 3279 + }, + { + "epoch": 3.8805087252292223, + "grad_norm": 1.4336597919464111, + "learning_rate": 9.000000000000001e-07, + "loss": 0.0317, + "step": 3280 + }, + { + "epoch": 3.8816918071576456, + "grad_norm": 1.3567018508911133, + "learning_rate": 8.987500000000001e-07, + "loss": 0.0323, + "step": 3281 + }, + { + "epoch": 3.8828748890860694, + "grad_norm": 1.3563851118087769, + "learning_rate": 8.975000000000001e-07, + "loss": 0.0322, + "step": 3282 + }, + { + "epoch": 3.8840579710144927, + "grad_norm": 1.3853124380111694, + "learning_rate": 8.9625e-07, + "loss": 0.0339, + "step": 3283 + }, + { + "epoch": 3.8852410529429164, + "grad_norm": 1.3448734283447266, + "learning_rate": 8.95e-07, + "loss": 0.0327, + "step": 3284 + }, + { + "epoch": 3.8864241348713398, + "grad_norm": 1.6741636991500854, + "learning_rate": 8.9375e-07, + "loss": 0.0291, + "step": 3285 + }, + { + "epoch": 3.8876072167997635, + "grad_norm": 1.6587629318237305, + "learning_rate": 8.925e-07, + "loss": 0.0328, + "step": 3286 + }, + { + "epoch": 3.888790298728187, + "grad_norm": 1.5547525882720947, + "learning_rate": 8.9125e-07, + "loss": 0.0342, + "step": 3287 + }, + { + "epoch": 3.8899733806566106, + "grad_norm": 1.6980674266815186, + "learning_rate": 8.900000000000001e-07, + "loss": 0.0354, + "step": 3288 + }, + { + "epoch": 3.891156462585034, + "grad_norm": 1.6172635555267334, + "learning_rate": 8.887500000000001e-07, + "loss": 0.0322, + "step": 3289 + }, + { + "epoch": 3.8923395445134576, + "grad_norm": 1.4632196426391602, + "learning_rate": 8.875000000000001e-07, + "loss": 0.0326, + "step": 3290 + }, + { + "epoch": 3.893522626441881, + "grad_norm": 1.0993475914001465, + "learning_rate": 8.8625e-07, + "loss": 0.0301, + "step": 3291 + }, + { + "epoch": 3.8947057083703047, + "grad_norm": 1.6714733839035034, + "learning_rate": 8.85e-07, + "loss": 0.0325, + "step": 3292 + }, + { + "epoch": 3.895888790298728, + "grad_norm": 1.8067665100097656, + "learning_rate": 8.8375e-07, + "loss": 0.0317, + "step": 3293 + }, + { + "epoch": 3.897071872227152, + "grad_norm": 2.0884475708007812, + "learning_rate": 8.825e-07, + "loss": 0.04, + "step": 3294 + }, + { + "epoch": 3.8982549541555755, + "grad_norm": 1.3442925214767456, + "learning_rate": 8.8125e-07, + "loss": 0.0331, + "step": 3295 + }, + { + "epoch": 3.899438036083999, + "grad_norm": 2.0974361896514893, + "learning_rate": 8.8e-07, + "loss": 0.0335, + "step": 3296 + }, + { + "epoch": 3.900621118012422, + "grad_norm": 1.6030317544937134, + "learning_rate": 8.787500000000001e-07, + "loss": 0.0285, + "step": 3297 + }, + { + "epoch": 3.901804199940846, + "grad_norm": 1.6683521270751953, + "learning_rate": 8.775000000000001e-07, + "loss": 0.0325, + "step": 3298 + }, + { + "epoch": 3.9029872818692697, + "grad_norm": 1.298296332359314, + "learning_rate": 8.7625e-07, + "loss": 0.0322, + "step": 3299 + }, + { + "epoch": 3.904170363797693, + "grad_norm": 1.7013070583343506, + "learning_rate": 8.75e-07, + "loss": 0.0355, + "step": 3300 + }, + { + "epoch": 3.9053534457261163, + "grad_norm": 1.1435459852218628, + "learning_rate": 8.7375e-07, + "loss": 0.0308, + "step": 3301 + }, + { + "epoch": 3.90653652765454, + "grad_norm": 1.6445759534835815, + "learning_rate": 8.725e-07, + "loss": 0.036, + "step": 3302 + }, + { + "epoch": 3.907719609582964, + "grad_norm": 1.8408029079437256, + "learning_rate": 8.7125e-07, + "loss": 0.0377, + "step": 3303 + }, + { + "epoch": 3.908902691511387, + "grad_norm": 1.2845250368118286, + "learning_rate": 8.7e-07, + "loss": 0.0328, + "step": 3304 + }, + { + "epoch": 3.9100857734398105, + "grad_norm": 1.442566990852356, + "learning_rate": 8.6875e-07, + "loss": 0.0306, + "step": 3305 + }, + { + "epoch": 3.911268855368234, + "grad_norm": 1.7673218250274658, + "learning_rate": 8.675000000000001e-07, + "loss": 0.0345, + "step": 3306 + }, + { + "epoch": 3.912451937296658, + "grad_norm": 1.5461399555206299, + "learning_rate": 8.6625e-07, + "loss": 0.0327, + "step": 3307 + }, + { + "epoch": 3.9136350192250813, + "grad_norm": 1.4908536672592163, + "learning_rate": 8.65e-07, + "loss": 0.0358, + "step": 3308 + }, + { + "epoch": 3.914818101153505, + "grad_norm": 1.6870843172073364, + "learning_rate": 8.6375e-07, + "loss": 0.0325, + "step": 3309 + }, + { + "epoch": 3.9160011830819283, + "grad_norm": 1.2814102172851562, + "learning_rate": 8.625e-07, + "loss": 0.0315, + "step": 3310 + }, + { + "epoch": 3.917184265010352, + "grad_norm": 1.477423906326294, + "learning_rate": 8.6125e-07, + "loss": 0.0269, + "step": 3311 + }, + { + "epoch": 3.9183673469387754, + "grad_norm": 1.483696460723877, + "learning_rate": 8.6e-07, + "loss": 0.0325, + "step": 3312 + }, + { + "epoch": 3.919550428867199, + "grad_norm": 1.4053646326065063, + "learning_rate": 8.587500000000001e-07, + "loss": 0.0299, + "step": 3313 + }, + { + "epoch": 3.9207335107956225, + "grad_norm": 1.4305837154388428, + "learning_rate": 8.575000000000002e-07, + "loss": 0.0336, + "step": 3314 + }, + { + "epoch": 3.9219165927240462, + "grad_norm": 1.4591296911239624, + "learning_rate": 8.562500000000002e-07, + "loss": 0.0279, + "step": 3315 + }, + { + "epoch": 3.9230996746524696, + "grad_norm": 1.5162149667739868, + "learning_rate": 8.550000000000002e-07, + "loss": 0.0308, + "step": 3316 + }, + { + "epoch": 3.9242827565808933, + "grad_norm": 1.6896600723266602, + "learning_rate": 8.537500000000001e-07, + "loss": 0.0295, + "step": 3317 + }, + { + "epoch": 3.9254658385093166, + "grad_norm": 1.4891828298568726, + "learning_rate": 8.525000000000001e-07, + "loss": 0.0386, + "step": 3318 + }, + { + "epoch": 3.9266489204377404, + "grad_norm": 1.7853292226791382, + "learning_rate": 8.512500000000001e-07, + "loss": 0.0378, + "step": 3319 + }, + { + "epoch": 3.9278320023661637, + "grad_norm": 1.6269445419311523, + "learning_rate": 8.500000000000001e-07, + "loss": 0.0312, + "step": 3320 + }, + { + "epoch": 3.9290150842945875, + "grad_norm": 1.467261552810669, + "learning_rate": 8.487500000000001e-07, + "loss": 0.0307, + "step": 3321 + }, + { + "epoch": 3.930198166223011, + "grad_norm": 1.7355353832244873, + "learning_rate": 8.475000000000001e-07, + "loss": 0.0367, + "step": 3322 + }, + { + "epoch": 3.9313812481514345, + "grad_norm": 1.4065600633621216, + "learning_rate": 8.462500000000002e-07, + "loss": 0.0335, + "step": 3323 + }, + { + "epoch": 3.932564330079858, + "grad_norm": 1.7136565446853638, + "learning_rate": 8.450000000000002e-07, + "loss": 0.0375, + "step": 3324 + }, + { + "epoch": 3.9337474120082816, + "grad_norm": 1.5231754779815674, + "learning_rate": 8.437500000000001e-07, + "loss": 0.0324, + "step": 3325 + }, + { + "epoch": 3.9349304939367054, + "grad_norm": 1.3040601015090942, + "learning_rate": 8.425000000000001e-07, + "loss": 0.0348, + "step": 3326 + }, + { + "epoch": 3.9361135758651287, + "grad_norm": 1.2653149366378784, + "learning_rate": 8.412500000000001e-07, + "loss": 0.0294, + "step": 3327 + }, + { + "epoch": 3.937296657793552, + "grad_norm": 1.4324527978897095, + "learning_rate": 8.400000000000001e-07, + "loss": 0.0363, + "step": 3328 + }, + { + "epoch": 3.9384797397219757, + "grad_norm": 1.4152510166168213, + "learning_rate": 8.387500000000001e-07, + "loss": 0.0333, + "step": 3329 + }, + { + "epoch": 3.9396628216503995, + "grad_norm": 1.5234920978546143, + "learning_rate": 8.375000000000001e-07, + "loss": 0.0304, + "step": 3330 + }, + { + "epoch": 3.940845903578823, + "grad_norm": 1.3950510025024414, + "learning_rate": 8.362500000000001e-07, + "loss": 0.0271, + "step": 3331 + }, + { + "epoch": 3.942028985507246, + "grad_norm": 1.3062162399291992, + "learning_rate": 8.350000000000002e-07, + "loss": 0.0317, + "step": 3332 + }, + { + "epoch": 3.94321206743567, + "grad_norm": 1.928748369216919, + "learning_rate": 8.337500000000001e-07, + "loss": 0.0314, + "step": 3333 + }, + { + "epoch": 3.9443951493640936, + "grad_norm": 1.4957391023635864, + "learning_rate": 8.325000000000001e-07, + "loss": 0.0346, + "step": 3334 + }, + { + "epoch": 3.945578231292517, + "grad_norm": 1.201930046081543, + "learning_rate": 8.312500000000001e-07, + "loss": 0.0297, + "step": 3335 + }, + { + "epoch": 3.9467613132209407, + "grad_norm": 1.3239091634750366, + "learning_rate": 8.300000000000001e-07, + "loss": 0.0289, + "step": 3336 + }, + { + "epoch": 3.947944395149364, + "grad_norm": 1.5993945598602295, + "learning_rate": 8.287500000000001e-07, + "loss": 0.0387, + "step": 3337 + }, + { + "epoch": 3.9491274770777878, + "grad_norm": 1.655487060546875, + "learning_rate": 8.275000000000001e-07, + "loss": 0.034, + "step": 3338 + }, + { + "epoch": 3.950310559006211, + "grad_norm": 1.4104193449020386, + "learning_rate": 8.262500000000001e-07, + "loss": 0.0332, + "step": 3339 + }, + { + "epoch": 3.951493640934635, + "grad_norm": 1.2167919874191284, + "learning_rate": 8.250000000000001e-07, + "loss": 0.0343, + "step": 3340 + }, + { + "epoch": 3.952676722863058, + "grad_norm": 1.3522578477859497, + "learning_rate": 8.237500000000001e-07, + "loss": 0.032, + "step": 3341 + }, + { + "epoch": 3.953859804791482, + "grad_norm": 1.454064130783081, + "learning_rate": 8.225000000000001e-07, + "loss": 0.0325, + "step": 3342 + }, + { + "epoch": 3.9550428867199052, + "grad_norm": 1.500585913658142, + "learning_rate": 8.212500000000001e-07, + "loss": 0.0366, + "step": 3343 + }, + { + "epoch": 3.956225968648329, + "grad_norm": 1.5090453624725342, + "learning_rate": 8.200000000000001e-07, + "loss": 0.0332, + "step": 3344 + }, + { + "epoch": 3.9574090505767523, + "grad_norm": 1.5809963941574097, + "learning_rate": 8.187500000000001e-07, + "loss": 0.0317, + "step": 3345 + }, + { + "epoch": 3.958592132505176, + "grad_norm": 1.621708869934082, + "learning_rate": 8.175000000000001e-07, + "loss": 0.0342, + "step": 3346 + }, + { + "epoch": 3.9597752144335994, + "grad_norm": 1.380223035812378, + "learning_rate": 8.162500000000001e-07, + "loss": 0.0314, + "step": 3347 + }, + { + "epoch": 3.960958296362023, + "grad_norm": 1.6043046712875366, + "learning_rate": 8.150000000000001e-07, + "loss": 0.0364, + "step": 3348 + }, + { + "epoch": 3.962141378290447, + "grad_norm": 1.7058428525924683, + "learning_rate": 8.1375e-07, + "loss": 0.035, + "step": 3349 + }, + { + "epoch": 3.96332446021887, + "grad_norm": 1.4469683170318604, + "learning_rate": 8.125000000000001e-07, + "loss": 0.0299, + "step": 3350 + }, + { + "epoch": 3.9645075421472935, + "grad_norm": 1.4945610761642456, + "learning_rate": 8.112500000000001e-07, + "loss": 0.0332, + "step": 3351 + }, + { + "epoch": 3.9656906240757173, + "grad_norm": 1.4792635440826416, + "learning_rate": 8.100000000000001e-07, + "loss": 0.0304, + "step": 3352 + }, + { + "epoch": 3.966873706004141, + "grad_norm": 1.0090372562408447, + "learning_rate": 8.087500000000001e-07, + "loss": 0.0316, + "step": 3353 + }, + { + "epoch": 3.9680567879325643, + "grad_norm": 1.1982927322387695, + "learning_rate": 8.075000000000001e-07, + "loss": 0.0288, + "step": 3354 + }, + { + "epoch": 3.9692398698609876, + "grad_norm": 1.233250379562378, + "learning_rate": 8.062500000000001e-07, + "loss": 0.0304, + "step": 3355 + }, + { + "epoch": 3.9704229517894114, + "grad_norm": 1.3439667224884033, + "learning_rate": 8.050000000000001e-07, + "loss": 0.0314, + "step": 3356 + }, + { + "epoch": 3.971606033717835, + "grad_norm": 1.610595941543579, + "learning_rate": 8.0375e-07, + "loss": 0.0297, + "step": 3357 + }, + { + "epoch": 3.9727891156462585, + "grad_norm": 1.3803462982177734, + "learning_rate": 8.025e-07, + "loss": 0.0325, + "step": 3358 + }, + { + "epoch": 3.973972197574682, + "grad_norm": 1.4965463876724243, + "learning_rate": 8.012500000000001e-07, + "loss": 0.0317, + "step": 3359 + }, + { + "epoch": 3.9751552795031055, + "grad_norm": 1.5316029787063599, + "learning_rate": 8.000000000000001e-07, + "loss": 0.0308, + "step": 3360 + }, + { + "epoch": 3.9763383614315293, + "grad_norm": 1.430654764175415, + "learning_rate": 7.987500000000001e-07, + "loss": 0.0325, + "step": 3361 + }, + { + "epoch": 3.9775214433599526, + "grad_norm": 1.4609973430633545, + "learning_rate": 7.975000000000001e-07, + "loss": 0.0354, + "step": 3362 + }, + { + "epoch": 3.9787045252883764, + "grad_norm": 1.407643437385559, + "learning_rate": 7.962500000000001e-07, + "loss": 0.0312, + "step": 3363 + }, + { + "epoch": 3.9798876072167997, + "grad_norm": 1.603326678276062, + "learning_rate": 7.950000000000001e-07, + "loss": 0.0314, + "step": 3364 + }, + { + "epoch": 3.9810706891452234, + "grad_norm": 1.4457813501358032, + "learning_rate": 7.9375e-07, + "loss": 0.0273, + "step": 3365 + }, + { + "epoch": 3.9822537710736468, + "grad_norm": 1.5081816911697388, + "learning_rate": 7.925e-07, + "loss": 0.0344, + "step": 3366 + }, + { + "epoch": 3.9834368530020705, + "grad_norm": 1.3513822555541992, + "learning_rate": 7.9125e-07, + "loss": 0.0283, + "step": 3367 + }, + { + "epoch": 3.984619934930494, + "grad_norm": 1.615708589553833, + "learning_rate": 7.900000000000001e-07, + "loss": 0.0346, + "step": 3368 + }, + { + "epoch": 3.9858030168589176, + "grad_norm": 1.466617465019226, + "learning_rate": 7.887500000000001e-07, + "loss": 0.0289, + "step": 3369 + }, + { + "epoch": 3.986986098787341, + "grad_norm": 1.378891110420227, + "learning_rate": 7.875000000000001e-07, + "loss": 0.034, + "step": 3370 + }, + { + "epoch": 3.9881691807157647, + "grad_norm": 1.8301976919174194, + "learning_rate": 7.862500000000001e-07, + "loss": 0.0332, + "step": 3371 + }, + { + "epoch": 3.989352262644188, + "grad_norm": 1.4099934101104736, + "learning_rate": 7.850000000000001e-07, + "loss": 0.033, + "step": 3372 + }, + { + "epoch": 3.9905353445726117, + "grad_norm": 1.3048890829086304, + "learning_rate": 7.8375e-07, + "loss": 0.0313, + "step": 3373 + }, + { + "epoch": 3.991718426501035, + "grad_norm": 1.4172672033309937, + "learning_rate": 7.825e-07, + "loss": 0.0321, + "step": 3374 + }, + { + "epoch": 3.992901508429459, + "grad_norm": 1.366847038269043, + "learning_rate": 7.8125e-07, + "loss": 0.034, + "step": 3375 + }, + { + "epoch": 3.9940845903578825, + "grad_norm": 1.7883164882659912, + "learning_rate": 7.8e-07, + "loss": 0.0337, + "step": 3376 + }, + { + "epoch": 3.995267672286306, + "grad_norm": 1.383488416671753, + "learning_rate": 7.787500000000001e-07, + "loss": 0.0312, + "step": 3377 + }, + { + "epoch": 3.996450754214729, + "grad_norm": 1.4037846326828003, + "learning_rate": 7.775000000000001e-07, + "loss": 0.0337, + "step": 3378 + }, + { + "epoch": 3.997633836143153, + "grad_norm": 1.6504777669906616, + "learning_rate": 7.762500000000001e-07, + "loss": 0.0291, + "step": 3379 + }, + { + "epoch": 3.9988169180715767, + "grad_norm": 1.2041751146316528, + "learning_rate": 7.750000000000001e-07, + "loss": 0.0309, + "step": 3380 + }, + { + "epoch": 4.0, + "grad_norm": 1.4494342803955078, + "learning_rate": 7.737500000000001e-07, + "loss": 0.0333, + "step": 3381 + }, + { + "epoch": 4.001183081928423, + "grad_norm": 0.6712071299552917, + "learning_rate": 7.725e-07, + "loss": 0.0137, + "step": 3382 + }, + { + "epoch": 4.0023661638568475, + "grad_norm": 0.9252065420150757, + "learning_rate": 7.7125e-07, + "loss": 0.0163, + "step": 3383 + }, + { + "epoch": 4.003549245785271, + "grad_norm": 0.7047213315963745, + "learning_rate": 7.7e-07, + "loss": 0.0134, + "step": 3384 + }, + { + "epoch": 4.004732327713694, + "grad_norm": 0.8338692784309387, + "learning_rate": 7.6875e-07, + "loss": 0.0148, + "step": 3385 + }, + { + "epoch": 4.0059154096421175, + "grad_norm": 0.8877047300338745, + "learning_rate": 7.675000000000001e-07, + "loss": 0.0151, + "step": 3386 + }, + { + "epoch": 4.007098491570542, + "grad_norm": 0.7846319675445557, + "learning_rate": 7.662500000000001e-07, + "loss": 0.0128, + "step": 3387 + }, + { + "epoch": 4.008281573498965, + "grad_norm": 0.9324353337287903, + "learning_rate": 7.650000000000001e-07, + "loss": 0.0135, + "step": 3388 + }, + { + "epoch": 4.009464655427388, + "grad_norm": 0.8601680397987366, + "learning_rate": 7.637500000000001e-07, + "loss": 0.0176, + "step": 3389 + }, + { + "epoch": 4.010647737355812, + "grad_norm": 0.919695258140564, + "learning_rate": 7.625e-07, + "loss": 0.0131, + "step": 3390 + }, + { + "epoch": 4.011830819284236, + "grad_norm": 1.0925676822662354, + "learning_rate": 7.6125e-07, + "loss": 0.0151, + "step": 3391 + }, + { + "epoch": 4.013013901212659, + "grad_norm": 0.9757835268974304, + "learning_rate": 7.6e-07, + "loss": 0.0147, + "step": 3392 + }, + { + "epoch": 4.014196983141082, + "grad_norm": 1.1667841672897339, + "learning_rate": 7.5875e-07, + "loss": 0.0138, + "step": 3393 + }, + { + "epoch": 4.015380065069506, + "grad_norm": 1.2077858448028564, + "learning_rate": 7.575000000000001e-07, + "loss": 0.0148, + "step": 3394 + }, + { + "epoch": 4.01656314699793, + "grad_norm": 1.3713231086730957, + "learning_rate": 7.562500000000001e-07, + "loss": 0.0146, + "step": 3395 + }, + { + "epoch": 4.017746228926353, + "grad_norm": 1.5427119731903076, + "learning_rate": 7.550000000000001e-07, + "loss": 0.0144, + "step": 3396 + }, + { + "epoch": 4.018929310854777, + "grad_norm": 1.093886137008667, + "learning_rate": 7.537500000000001e-07, + "loss": 0.0136, + "step": 3397 + }, + { + "epoch": 4.0201123927832, + "grad_norm": 1.303457260131836, + "learning_rate": 7.525e-07, + "loss": 0.0162, + "step": 3398 + }, + { + "epoch": 4.021295474711624, + "grad_norm": 0.9375545382499695, + "learning_rate": 7.5125e-07, + "loss": 0.0154, + "step": 3399 + }, + { + "epoch": 4.022478556640047, + "grad_norm": 1.1201436519622803, + "learning_rate": 7.5e-07, + "loss": 0.0147, + "step": 3400 + }, + { + "epoch": 4.023661638568471, + "grad_norm": 1.72199285030365, + "learning_rate": 7.4875e-07, + "loss": 0.0157, + "step": 3401 + }, + { + "epoch": 4.024844720496894, + "grad_norm": 0.8972126245498657, + "learning_rate": 7.475e-07, + "loss": 0.0142, + "step": 3402 + }, + { + "epoch": 4.026027802425318, + "grad_norm": 1.4768855571746826, + "learning_rate": 7.462500000000001e-07, + "loss": 0.015, + "step": 3403 + }, + { + "epoch": 4.0272108843537415, + "grad_norm": 1.3319556713104248, + "learning_rate": 7.450000000000001e-07, + "loss": 0.0177, + "step": 3404 + }, + { + "epoch": 4.028393966282165, + "grad_norm": 1.133811116218567, + "learning_rate": 7.437500000000001e-07, + "loss": 0.0146, + "step": 3405 + }, + { + "epoch": 4.029577048210588, + "grad_norm": 1.1574431657791138, + "learning_rate": 7.425e-07, + "loss": 0.0177, + "step": 3406 + }, + { + "epoch": 4.030760130139012, + "grad_norm": 1.2803245782852173, + "learning_rate": 7.4125e-07, + "loss": 0.0156, + "step": 3407 + }, + { + "epoch": 4.031943212067436, + "grad_norm": 1.2277494668960571, + "learning_rate": 7.4e-07, + "loss": 0.0144, + "step": 3408 + }, + { + "epoch": 4.033126293995859, + "grad_norm": 1.1839910745620728, + "learning_rate": 7.3875e-07, + "loss": 0.0153, + "step": 3409 + }, + { + "epoch": 4.034309375924283, + "grad_norm": 1.3643828630447388, + "learning_rate": 7.375e-07, + "loss": 0.0145, + "step": 3410 + }, + { + "epoch": 4.0354924578527065, + "grad_norm": 1.0742640495300293, + "learning_rate": 7.3625e-07, + "loss": 0.0126, + "step": 3411 + }, + { + "epoch": 4.03667553978113, + "grad_norm": 1.3911257982254028, + "learning_rate": 7.350000000000001e-07, + "loss": 0.0179, + "step": 3412 + }, + { + "epoch": 4.037858621709553, + "grad_norm": 0.8934176564216614, + "learning_rate": 7.337500000000001e-07, + "loss": 0.0153, + "step": 3413 + }, + { + "epoch": 4.039041703637977, + "grad_norm": 1.0038830041885376, + "learning_rate": 7.325e-07, + "loss": 0.0162, + "step": 3414 + }, + { + "epoch": 4.040224785566401, + "grad_norm": 1.2437002658843994, + "learning_rate": 7.3125e-07, + "loss": 0.0155, + "step": 3415 + }, + { + "epoch": 4.041407867494824, + "grad_norm": 0.9356801509857178, + "learning_rate": 7.3e-07, + "loss": 0.0138, + "step": 3416 + }, + { + "epoch": 4.042590949423247, + "grad_norm": 0.885857880115509, + "learning_rate": 7.2875e-07, + "loss": 0.0148, + "step": 3417 + }, + { + "epoch": 4.0437740313516715, + "grad_norm": 1.29246985912323, + "learning_rate": 7.275e-07, + "loss": 0.0161, + "step": 3418 + }, + { + "epoch": 4.044957113280095, + "grad_norm": 1.0270977020263672, + "learning_rate": 7.2625e-07, + "loss": 0.0136, + "step": 3419 + }, + { + "epoch": 4.046140195208518, + "grad_norm": 1.4631562232971191, + "learning_rate": 7.25e-07, + "loss": 0.0164, + "step": 3420 + }, + { + "epoch": 4.047323277136941, + "grad_norm": 1.3347618579864502, + "learning_rate": 7.237500000000001e-07, + "loss": 0.0157, + "step": 3421 + }, + { + "epoch": 4.048506359065366, + "grad_norm": 1.1093395948410034, + "learning_rate": 7.225e-07, + "loss": 0.0162, + "step": 3422 + }, + { + "epoch": 4.049689440993789, + "grad_norm": 0.9512934684753418, + "learning_rate": 7.2125e-07, + "loss": 0.0163, + "step": 3423 + }, + { + "epoch": 4.050872522922212, + "grad_norm": 0.8720839023590088, + "learning_rate": 7.2e-07, + "loss": 0.0154, + "step": 3424 + }, + { + "epoch": 4.0520556048506355, + "grad_norm": 1.012698769569397, + "learning_rate": 7.1875e-07, + "loss": 0.0131, + "step": 3425 + }, + { + "epoch": 4.05323868677906, + "grad_norm": 1.0069795846939087, + "learning_rate": 7.175e-07, + "loss": 0.0145, + "step": 3426 + }, + { + "epoch": 4.054421768707483, + "grad_norm": 1.3598345518112183, + "learning_rate": 7.1625e-07, + "loss": 0.0144, + "step": 3427 + }, + { + "epoch": 4.055604850635906, + "grad_norm": 1.2950471639633179, + "learning_rate": 7.15e-07, + "loss": 0.0167, + "step": 3428 + }, + { + "epoch": 4.05678793256433, + "grad_norm": 1.27305006980896, + "learning_rate": 7.1375e-07, + "loss": 0.0159, + "step": 3429 + }, + { + "epoch": 4.057971014492754, + "grad_norm": 0.9707962274551392, + "learning_rate": 7.125e-07, + "loss": 0.0121, + "step": 3430 + }, + { + "epoch": 4.059154096421177, + "grad_norm": 0.8594356775283813, + "learning_rate": 7.1125e-07, + "loss": 0.0124, + "step": 3431 + }, + { + "epoch": 4.0603371783496005, + "grad_norm": 1.2895547151565552, + "learning_rate": 7.1e-07, + "loss": 0.0164, + "step": 3432 + }, + { + "epoch": 4.061520260278024, + "grad_norm": 1.5395681858062744, + "learning_rate": 7.0875e-07, + "loss": 0.016, + "step": 3433 + }, + { + "epoch": 4.062703342206448, + "grad_norm": 1.3186559677124023, + "learning_rate": 7.075e-07, + "loss": 0.0151, + "step": 3434 + }, + { + "epoch": 4.063886424134871, + "grad_norm": 1.0605987310409546, + "learning_rate": 7.0625e-07, + "loss": 0.0145, + "step": 3435 + }, + { + "epoch": 4.065069506063295, + "grad_norm": 1.0249269008636475, + "learning_rate": 7.05e-07, + "loss": 0.0174, + "step": 3436 + }, + { + "epoch": 4.066252587991719, + "grad_norm": 0.9716159105300903, + "learning_rate": 7.0375e-07, + "loss": 0.0181, + "step": 3437 + }, + { + "epoch": 4.067435669920142, + "grad_norm": 1.3415510654449463, + "learning_rate": 7.025000000000002e-07, + "loss": 0.0137, + "step": 3438 + }, + { + "epoch": 4.0686187518485655, + "grad_norm": 1.0976347923278809, + "learning_rate": 7.012500000000002e-07, + "loss": 0.0144, + "step": 3439 + }, + { + "epoch": 4.069801833776989, + "grad_norm": 1.324601650238037, + "learning_rate": 7.000000000000001e-07, + "loss": 0.0138, + "step": 3440 + }, + { + "epoch": 4.070984915705413, + "grad_norm": 1.328048825263977, + "learning_rate": 6.987500000000001e-07, + "loss": 0.0182, + "step": 3441 + }, + { + "epoch": 4.072167997633836, + "grad_norm": 1.115700125694275, + "learning_rate": 6.975000000000001e-07, + "loss": 0.0119, + "step": 3442 + }, + { + "epoch": 4.07335107956226, + "grad_norm": 1.0370335578918457, + "learning_rate": 6.962500000000001e-07, + "loss": 0.0172, + "step": 3443 + }, + { + "epoch": 4.074534161490683, + "grad_norm": 1.1017335653305054, + "learning_rate": 6.950000000000001e-07, + "loss": 0.0136, + "step": 3444 + }, + { + "epoch": 4.075717243419107, + "grad_norm": 1.8941309452056885, + "learning_rate": 6.937500000000001e-07, + "loss": 0.0175, + "step": 3445 + }, + { + "epoch": 4.07690032534753, + "grad_norm": 0.8066739439964294, + "learning_rate": 6.925000000000001e-07, + "loss": 0.0151, + "step": 3446 + }, + { + "epoch": 4.078083407275954, + "grad_norm": 1.0229219198226929, + "learning_rate": 6.912500000000002e-07, + "loss": 0.0163, + "step": 3447 + }, + { + "epoch": 4.079266489204377, + "grad_norm": 0.9591362476348877, + "learning_rate": 6.900000000000001e-07, + "loss": 0.0144, + "step": 3448 + }, + { + "epoch": 4.080449571132801, + "grad_norm": 1.1412920951843262, + "learning_rate": 6.887500000000001e-07, + "loss": 0.017, + "step": 3449 + }, + { + "epoch": 4.081632653061225, + "grad_norm": 1.2124912738800049, + "learning_rate": 6.875000000000001e-07, + "loss": 0.0188, + "step": 3450 + }, + { + "epoch": 4.082815734989648, + "grad_norm": 1.1085692644119263, + "learning_rate": 6.862500000000001e-07, + "loss": 0.0134, + "step": 3451 + }, + { + "epoch": 4.083998816918071, + "grad_norm": 1.1501420736312866, + "learning_rate": 6.850000000000001e-07, + "loss": 0.0153, + "step": 3452 + }, + { + "epoch": 4.085181898846495, + "grad_norm": 1.647203803062439, + "learning_rate": 6.837500000000001e-07, + "loss": 0.0199, + "step": 3453 + }, + { + "epoch": 4.086364980774919, + "grad_norm": 1.1346039772033691, + "learning_rate": 6.825000000000001e-07, + "loss": 0.0157, + "step": 3454 + }, + { + "epoch": 4.087548062703342, + "grad_norm": 1.624648928642273, + "learning_rate": 6.812500000000001e-07, + "loss": 0.0168, + "step": 3455 + }, + { + "epoch": 4.088731144631765, + "grad_norm": 1.269086480140686, + "learning_rate": 6.800000000000001e-07, + "loss": 0.0148, + "step": 3456 + }, + { + "epoch": 4.0899142265601895, + "grad_norm": 1.434822678565979, + "learning_rate": 6.787500000000001e-07, + "loss": 0.0177, + "step": 3457 + }, + { + "epoch": 4.091097308488613, + "grad_norm": 1.0883805751800537, + "learning_rate": 6.775000000000001e-07, + "loss": 0.0155, + "step": 3458 + }, + { + "epoch": 4.092280390417036, + "grad_norm": 1.2063273191452026, + "learning_rate": 6.762500000000001e-07, + "loss": 0.0136, + "step": 3459 + }, + { + "epoch": 4.0934634723454595, + "grad_norm": 1.3631967306137085, + "learning_rate": 6.750000000000001e-07, + "loss": 0.0135, + "step": 3460 + }, + { + "epoch": 4.094646554273884, + "grad_norm": 1.3141480684280396, + "learning_rate": 6.737500000000001e-07, + "loss": 0.0168, + "step": 3461 + }, + { + "epoch": 4.095829636202307, + "grad_norm": 1.5140384435653687, + "learning_rate": 6.725000000000001e-07, + "loss": 0.015, + "step": 3462 + }, + { + "epoch": 4.09701271813073, + "grad_norm": 1.0243693590164185, + "learning_rate": 6.712500000000001e-07, + "loss": 0.014, + "step": 3463 + }, + { + "epoch": 4.0981958000591545, + "grad_norm": 1.343532919883728, + "learning_rate": 6.7e-07, + "loss": 0.0141, + "step": 3464 + }, + { + "epoch": 4.099378881987578, + "grad_norm": 1.693604826927185, + "learning_rate": 6.687500000000001e-07, + "loss": 0.0139, + "step": 3465 + }, + { + "epoch": 4.100561963916001, + "grad_norm": 1.25705885887146, + "learning_rate": 6.675000000000001e-07, + "loss": 0.0145, + "step": 3466 + }, + { + "epoch": 4.1017450458444245, + "grad_norm": 1.3815205097198486, + "learning_rate": 6.662500000000001e-07, + "loss": 0.0144, + "step": 3467 + }, + { + "epoch": 4.102928127772849, + "grad_norm": 1.5639536380767822, + "learning_rate": 6.650000000000001e-07, + "loss": 0.0178, + "step": 3468 + }, + { + "epoch": 4.104111209701272, + "grad_norm": 1.2690300941467285, + "learning_rate": 6.637500000000001e-07, + "loss": 0.0117, + "step": 3469 + }, + { + "epoch": 4.105294291629695, + "grad_norm": 0.9780581593513489, + "learning_rate": 6.625000000000001e-07, + "loss": 0.0147, + "step": 3470 + }, + { + "epoch": 4.106477373558119, + "grad_norm": 0.9527963995933533, + "learning_rate": 6.612500000000001e-07, + "loss": 0.0164, + "step": 3471 + }, + { + "epoch": 4.107660455486543, + "grad_norm": 1.3045427799224854, + "learning_rate": 6.6e-07, + "loss": 0.0155, + "step": 3472 + }, + { + "epoch": 4.108843537414966, + "grad_norm": 1.2041444778442383, + "learning_rate": 6.5875e-07, + "loss": 0.0143, + "step": 3473 + }, + { + "epoch": 4.110026619343389, + "grad_norm": 0.9065019488334656, + "learning_rate": 6.575000000000001e-07, + "loss": 0.012, + "step": 3474 + }, + { + "epoch": 4.111209701271813, + "grad_norm": 1.2192661762237549, + "learning_rate": 6.562500000000001e-07, + "loss": 0.015, + "step": 3475 + }, + { + "epoch": 4.112392783200237, + "grad_norm": 0.9799237251281738, + "learning_rate": 6.550000000000001e-07, + "loss": 0.0141, + "step": 3476 + }, + { + "epoch": 4.11357586512866, + "grad_norm": 1.1047767400741577, + "learning_rate": 6.537500000000001e-07, + "loss": 0.015, + "step": 3477 + }, + { + "epoch": 4.114758947057084, + "grad_norm": 1.0832250118255615, + "learning_rate": 6.525000000000001e-07, + "loss": 0.0153, + "step": 3478 + }, + { + "epoch": 4.115942028985507, + "grad_norm": 1.4630954265594482, + "learning_rate": 6.512500000000001e-07, + "loss": 0.0156, + "step": 3479 + }, + { + "epoch": 4.117125110913931, + "grad_norm": 1.2095292806625366, + "learning_rate": 6.5e-07, + "loss": 0.0146, + "step": 3480 + }, + { + "epoch": 4.118308192842354, + "grad_norm": 1.199334740638733, + "learning_rate": 6.4875e-07, + "loss": 0.0133, + "step": 3481 + }, + { + "epoch": 4.119491274770778, + "grad_norm": 1.2467288970947266, + "learning_rate": 6.475e-07, + "loss": 0.0129, + "step": 3482 + }, + { + "epoch": 4.120674356699201, + "grad_norm": 1.3035966157913208, + "learning_rate": 6.462500000000001e-07, + "loss": 0.015, + "step": 3483 + }, + { + "epoch": 4.121857438627625, + "grad_norm": 1.2883275747299194, + "learning_rate": 6.450000000000001e-07, + "loss": 0.0153, + "step": 3484 + }, + { + "epoch": 4.1230405205560485, + "grad_norm": 1.1759676933288574, + "learning_rate": 6.437500000000001e-07, + "loss": 0.0154, + "step": 3485 + }, + { + "epoch": 4.124223602484472, + "grad_norm": 1.1833043098449707, + "learning_rate": 6.425000000000001e-07, + "loss": 0.0136, + "step": 3486 + }, + { + "epoch": 4.125406684412896, + "grad_norm": 1.0531272888183594, + "learning_rate": 6.412500000000001e-07, + "loss": 0.0143, + "step": 3487 + }, + { + "epoch": 4.126589766341319, + "grad_norm": 1.3032867908477783, + "learning_rate": 6.4e-07, + "loss": 0.0158, + "step": 3488 + }, + { + "epoch": 4.127772848269743, + "grad_norm": 1.0476019382476807, + "learning_rate": 6.3875e-07, + "loss": 0.0129, + "step": 3489 + }, + { + "epoch": 4.128955930198166, + "grad_norm": 1.4543040990829468, + "learning_rate": 6.375e-07, + "loss": 0.0175, + "step": 3490 + }, + { + "epoch": 4.13013901212659, + "grad_norm": 1.7004433870315552, + "learning_rate": 6.3625e-07, + "loss": 0.0172, + "step": 3491 + }, + { + "epoch": 4.1313220940550135, + "grad_norm": 1.0993958711624146, + "learning_rate": 6.350000000000001e-07, + "loss": 0.0134, + "step": 3492 + }, + { + "epoch": 4.132505175983437, + "grad_norm": 1.359779953956604, + "learning_rate": 6.337500000000001e-07, + "loss": 0.0158, + "step": 3493 + }, + { + "epoch": 4.13368825791186, + "grad_norm": 1.2762892246246338, + "learning_rate": 6.325000000000001e-07, + "loss": 0.0167, + "step": 3494 + }, + { + "epoch": 4.134871339840284, + "grad_norm": 1.3144510984420776, + "learning_rate": 6.312500000000001e-07, + "loss": 0.0144, + "step": 3495 + }, + { + "epoch": 4.136054421768708, + "grad_norm": 1.239758014678955, + "learning_rate": 6.3e-07, + "loss": 0.0157, + "step": 3496 + }, + { + "epoch": 4.137237503697131, + "grad_norm": 0.9998862147331238, + "learning_rate": 6.2875e-07, + "loss": 0.0135, + "step": 3497 + }, + { + "epoch": 4.138420585625554, + "grad_norm": 0.8762887120246887, + "learning_rate": 6.275e-07, + "loss": 0.0123, + "step": 3498 + }, + { + "epoch": 4.1396036675539785, + "grad_norm": 1.1912308931350708, + "learning_rate": 6.2625e-07, + "loss": 0.016, + "step": 3499 + }, + { + "epoch": 4.140786749482402, + "grad_norm": 0.6407239437103271, + "learning_rate": 6.25e-07, + "loss": 0.0139, + "step": 3500 + }, + { + "epoch": 4.141969831410825, + "grad_norm": 1.2361286878585815, + "learning_rate": 6.237500000000001e-07, + "loss": 0.0163, + "step": 3501 + }, + { + "epoch": 4.143152913339248, + "grad_norm": 1.0342464447021484, + "learning_rate": 6.225000000000001e-07, + "loss": 0.0133, + "step": 3502 + }, + { + "epoch": 4.144335995267673, + "grad_norm": 1.0916259288787842, + "learning_rate": 6.212500000000001e-07, + "loss": 0.0168, + "step": 3503 + }, + { + "epoch": 4.145519077196096, + "grad_norm": 1.378536581993103, + "learning_rate": 6.200000000000001e-07, + "loss": 0.0164, + "step": 3504 + }, + { + "epoch": 4.146702159124519, + "grad_norm": 1.0369548797607422, + "learning_rate": 6.1875e-07, + "loss": 0.0139, + "step": 3505 + }, + { + "epoch": 4.1478852410529425, + "grad_norm": 1.2113922834396362, + "learning_rate": 6.175e-07, + "loss": 0.0159, + "step": 3506 + }, + { + "epoch": 4.149068322981367, + "grad_norm": 1.327402114868164, + "learning_rate": 6.1625e-07, + "loss": 0.0135, + "step": 3507 + }, + { + "epoch": 4.15025140490979, + "grad_norm": 1.068105697631836, + "learning_rate": 6.15e-07, + "loss": 0.0146, + "step": 3508 + }, + { + "epoch": 4.151434486838213, + "grad_norm": 1.0456892251968384, + "learning_rate": 6.137500000000001e-07, + "loss": 0.0146, + "step": 3509 + }, + { + "epoch": 4.152617568766637, + "grad_norm": 1.563988208770752, + "learning_rate": 6.125000000000001e-07, + "loss": 0.0164, + "step": 3510 + }, + { + "epoch": 4.153800650695061, + "grad_norm": 1.2215781211853027, + "learning_rate": 6.112500000000001e-07, + "loss": 0.0169, + "step": 3511 + }, + { + "epoch": 4.154983732623484, + "grad_norm": 1.1518625020980835, + "learning_rate": 6.100000000000001e-07, + "loss": 0.0192, + "step": 3512 + }, + { + "epoch": 4.1561668145519075, + "grad_norm": 1.061421275138855, + "learning_rate": 6.0875e-07, + "loss": 0.0115, + "step": 3513 + }, + { + "epoch": 4.157349896480332, + "grad_norm": 1.1260621547698975, + "learning_rate": 6.075e-07, + "loss": 0.0117, + "step": 3514 + }, + { + "epoch": 4.158532978408755, + "grad_norm": 1.472580909729004, + "learning_rate": 6.0625e-07, + "loss": 0.0161, + "step": 3515 + }, + { + "epoch": 4.159716060337178, + "grad_norm": 1.0845364332199097, + "learning_rate": 6.05e-07, + "loss": 0.0136, + "step": 3516 + }, + { + "epoch": 4.160899142265602, + "grad_norm": 1.1387759447097778, + "learning_rate": 6.0375e-07, + "loss": 0.0156, + "step": 3517 + }, + { + "epoch": 4.162082224194026, + "grad_norm": 1.366635799407959, + "learning_rate": 6.025000000000001e-07, + "loss": 0.0141, + "step": 3518 + }, + { + "epoch": 4.163265306122449, + "grad_norm": 1.1506577730178833, + "learning_rate": 6.012500000000001e-07, + "loss": 0.0138, + "step": 3519 + }, + { + "epoch": 4.1644483880508725, + "grad_norm": 1.2088648080825806, + "learning_rate": 6.000000000000001e-07, + "loss": 0.0158, + "step": 3520 + }, + { + "epoch": 4.165631469979296, + "grad_norm": 0.9968836903572083, + "learning_rate": 5.9875e-07, + "loss": 0.017, + "step": 3521 + }, + { + "epoch": 4.16681455190772, + "grad_norm": 1.1911300420761108, + "learning_rate": 5.975e-07, + "loss": 0.0139, + "step": 3522 + }, + { + "epoch": 4.167997633836143, + "grad_norm": 1.258027195930481, + "learning_rate": 5.9625e-07, + "loss": 0.0144, + "step": 3523 + }, + { + "epoch": 4.169180715764567, + "grad_norm": 1.3163537979125977, + "learning_rate": 5.95e-07, + "loss": 0.0155, + "step": 3524 + }, + { + "epoch": 4.17036379769299, + "grad_norm": 1.270803451538086, + "learning_rate": 5.9375e-07, + "loss": 0.0126, + "step": 3525 + }, + { + "epoch": 4.171546879621414, + "grad_norm": 1.5761808156967163, + "learning_rate": 5.925e-07, + "loss": 0.0146, + "step": 3526 + }, + { + "epoch": 4.172729961549837, + "grad_norm": 1.1174696683883667, + "learning_rate": 5.912500000000001e-07, + "loss": 0.0139, + "step": 3527 + }, + { + "epoch": 4.173913043478261, + "grad_norm": 1.3147093057632446, + "learning_rate": 5.900000000000001e-07, + "loss": 0.0134, + "step": 3528 + }, + { + "epoch": 4.175096125406684, + "grad_norm": 1.1345754861831665, + "learning_rate": 5.8875e-07, + "loss": 0.0124, + "step": 3529 + }, + { + "epoch": 4.176279207335108, + "grad_norm": 1.1139323711395264, + "learning_rate": 5.875e-07, + "loss": 0.0157, + "step": 3530 + }, + { + "epoch": 4.177462289263532, + "grad_norm": 1.1677659749984741, + "learning_rate": 5.8625e-07, + "loss": 0.0145, + "step": 3531 + }, + { + "epoch": 4.178645371191955, + "grad_norm": 1.1955451965332031, + "learning_rate": 5.850000000000001e-07, + "loss": 0.0116, + "step": 3532 + }, + { + "epoch": 4.179828453120378, + "grad_norm": 1.0077284574508667, + "learning_rate": 5.837500000000001e-07, + "loss": 0.0132, + "step": 3533 + }, + { + "epoch": 4.181011535048802, + "grad_norm": 0.9232893586158752, + "learning_rate": 5.825e-07, + "loss": 0.0157, + "step": 3534 + }, + { + "epoch": 4.182194616977226, + "grad_norm": 1.0326405763626099, + "learning_rate": 5.8125e-07, + "loss": 0.0141, + "step": 3535 + }, + { + "epoch": 4.183377698905649, + "grad_norm": 1.4907057285308838, + "learning_rate": 5.800000000000001e-07, + "loss": 0.0145, + "step": 3536 + }, + { + "epoch": 4.184560780834072, + "grad_norm": 0.8111011385917664, + "learning_rate": 5.787500000000001e-07, + "loss": 0.0117, + "step": 3537 + }, + { + "epoch": 4.1857438627624965, + "grad_norm": 0.8658562302589417, + "learning_rate": 5.775000000000001e-07, + "loss": 0.0126, + "step": 3538 + }, + { + "epoch": 4.18692694469092, + "grad_norm": 1.4250104427337646, + "learning_rate": 5.762500000000001e-07, + "loss": 0.0157, + "step": 3539 + }, + { + "epoch": 4.188110026619343, + "grad_norm": 1.5428072214126587, + "learning_rate": 5.750000000000001e-07, + "loss": 0.0157, + "step": 3540 + }, + { + "epoch": 4.189293108547767, + "grad_norm": 1.473046898841858, + "learning_rate": 5.737500000000001e-07, + "loss": 0.0163, + "step": 3541 + }, + { + "epoch": 4.190476190476191, + "grad_norm": 1.80694580078125, + "learning_rate": 5.725e-07, + "loss": 0.0154, + "step": 3542 + }, + { + "epoch": 4.191659272404614, + "grad_norm": 0.9662620425224304, + "learning_rate": 5.7125e-07, + "loss": 0.0154, + "step": 3543 + }, + { + "epoch": 4.192842354333037, + "grad_norm": 0.9812756776809692, + "learning_rate": 5.7e-07, + "loss": 0.0153, + "step": 3544 + }, + { + "epoch": 4.1940254362614615, + "grad_norm": 1.009981632232666, + "learning_rate": 5.687500000000001e-07, + "loss": 0.0123, + "step": 3545 + }, + { + "epoch": 4.195208518189885, + "grad_norm": 0.9284949898719788, + "learning_rate": 5.675000000000001e-07, + "loss": 0.0131, + "step": 3546 + }, + { + "epoch": 4.196391600118308, + "grad_norm": 1.0809892416000366, + "learning_rate": 5.662500000000001e-07, + "loss": 0.0167, + "step": 3547 + }, + { + "epoch": 4.1975746820467315, + "grad_norm": 1.3385117053985596, + "learning_rate": 5.650000000000001e-07, + "loss": 0.0141, + "step": 3548 + }, + { + "epoch": 4.198757763975156, + "grad_norm": 1.3884369134902954, + "learning_rate": 5.637500000000001e-07, + "loss": 0.0148, + "step": 3549 + }, + { + "epoch": 4.199940845903579, + "grad_norm": 2.714723587036133, + "learning_rate": 5.625e-07, + "loss": 0.0145, + "step": 3550 + }, + { + "epoch": 4.201123927832002, + "grad_norm": 1.371387004852295, + "learning_rate": 5.6125e-07, + "loss": 0.0159, + "step": 3551 + }, + { + "epoch": 4.202307009760426, + "grad_norm": 1.93416428565979, + "learning_rate": 5.6e-07, + "loss": 0.0169, + "step": 3552 + }, + { + "epoch": 4.20349009168885, + "grad_norm": 1.0045504570007324, + "learning_rate": 5.5875e-07, + "loss": 0.0131, + "step": 3553 + }, + { + "epoch": 4.204673173617273, + "grad_norm": 1.302680253982544, + "learning_rate": 5.575000000000001e-07, + "loss": 0.0167, + "step": 3554 + }, + { + "epoch": 4.205856255545696, + "grad_norm": 1.2461503744125366, + "learning_rate": 5.562500000000001e-07, + "loss": 0.0165, + "step": 3555 + }, + { + "epoch": 4.20703933747412, + "grad_norm": 1.1212751865386963, + "learning_rate": 5.550000000000001e-07, + "loss": 0.0135, + "step": 3556 + }, + { + "epoch": 4.208222419402544, + "grad_norm": 1.0680972337722778, + "learning_rate": 5.537500000000001e-07, + "loss": 0.0134, + "step": 3557 + }, + { + "epoch": 4.209405501330967, + "grad_norm": 1.0067449808120728, + "learning_rate": 5.525e-07, + "loss": 0.0176, + "step": 3558 + }, + { + "epoch": 4.210588583259391, + "grad_norm": 1.387519359588623, + "learning_rate": 5.5125e-07, + "loss": 0.0138, + "step": 3559 + }, + { + "epoch": 4.211771665187814, + "grad_norm": 1.4073983430862427, + "learning_rate": 5.5e-07, + "loss": 0.0127, + "step": 3560 + }, + { + "epoch": 4.212954747116238, + "grad_norm": 1.2140369415283203, + "learning_rate": 5.4875e-07, + "loss": 0.0126, + "step": 3561 + }, + { + "epoch": 4.214137829044661, + "grad_norm": 1.366350769996643, + "learning_rate": 5.475e-07, + "loss": 0.0157, + "step": 3562 + }, + { + "epoch": 4.215320910973085, + "grad_norm": 1.283484697341919, + "learning_rate": 5.462500000000001e-07, + "loss": 0.0156, + "step": 3563 + }, + { + "epoch": 4.216503992901508, + "grad_norm": 1.1450811624526978, + "learning_rate": 5.450000000000001e-07, + "loss": 0.0153, + "step": 3564 + }, + { + "epoch": 4.217687074829932, + "grad_norm": 1.4765201807022095, + "learning_rate": 5.437500000000001e-07, + "loss": 0.014, + "step": 3565 + }, + { + "epoch": 4.2188701567583555, + "grad_norm": 1.1871325969696045, + "learning_rate": 5.425e-07, + "loss": 0.0155, + "step": 3566 + }, + { + "epoch": 4.220053238686779, + "grad_norm": 1.2608391046524048, + "learning_rate": 5.4125e-07, + "loss": 0.0124, + "step": 3567 + }, + { + "epoch": 4.221236320615203, + "grad_norm": 1.1357945203781128, + "learning_rate": 5.4e-07, + "loss": 0.016, + "step": 3568 + }, + { + "epoch": 4.222419402543626, + "grad_norm": 1.1687713861465454, + "learning_rate": 5.3875e-07, + "loss": 0.0165, + "step": 3569 + }, + { + "epoch": 4.22360248447205, + "grad_norm": 1.1227748394012451, + "learning_rate": 5.375e-07, + "loss": 0.014, + "step": 3570 + }, + { + "epoch": 4.224785566400473, + "grad_norm": 1.002967357635498, + "learning_rate": 5.362500000000001e-07, + "loss": 0.0151, + "step": 3571 + }, + { + "epoch": 4.225968648328897, + "grad_norm": 1.3030284643173218, + "learning_rate": 5.350000000000001e-07, + "loss": 0.0141, + "step": 3572 + }, + { + "epoch": 4.2271517302573205, + "grad_norm": 1.3868446350097656, + "learning_rate": 5.337500000000001e-07, + "loss": 0.0154, + "step": 3573 + }, + { + "epoch": 4.228334812185744, + "grad_norm": 1.3573379516601562, + "learning_rate": 5.325e-07, + "loss": 0.0119, + "step": 3574 + }, + { + "epoch": 4.229517894114167, + "grad_norm": 0.966695249080658, + "learning_rate": 5.3125e-07, + "loss": 0.0154, + "step": 3575 + }, + { + "epoch": 4.230700976042591, + "grad_norm": 1.3644839525222778, + "learning_rate": 5.3e-07, + "loss": 0.0158, + "step": 3576 + }, + { + "epoch": 4.231884057971015, + "grad_norm": 0.989528477191925, + "learning_rate": 5.2875e-07, + "loss": 0.0143, + "step": 3577 + }, + { + "epoch": 4.233067139899438, + "grad_norm": 1.2114574909210205, + "learning_rate": 5.275e-07, + "loss": 0.0143, + "step": 3578 + }, + { + "epoch": 4.234250221827861, + "grad_norm": 1.3287700414657593, + "learning_rate": 5.2625e-07, + "loss": 0.0148, + "step": 3579 + }, + { + "epoch": 4.2354333037562855, + "grad_norm": 0.9921172857284546, + "learning_rate": 5.250000000000001e-07, + "loss": 0.0128, + "step": 3580 + }, + { + "epoch": 4.236616385684709, + "grad_norm": 1.3009345531463623, + "learning_rate": 5.237500000000001e-07, + "loss": 0.016, + "step": 3581 + }, + { + "epoch": 4.237799467613132, + "grad_norm": 1.3265211582183838, + "learning_rate": 5.225e-07, + "loss": 0.0156, + "step": 3582 + }, + { + "epoch": 4.238982549541555, + "grad_norm": 1.1114667654037476, + "learning_rate": 5.2125e-07, + "loss": 0.0129, + "step": 3583 + }, + { + "epoch": 4.24016563146998, + "grad_norm": 1.597588300704956, + "learning_rate": 5.2e-07, + "loss": 0.0155, + "step": 3584 + }, + { + "epoch": 4.241348713398403, + "grad_norm": 1.3502302169799805, + "learning_rate": 5.1875e-07, + "loss": 0.0144, + "step": 3585 + }, + { + "epoch": 4.242531795326826, + "grad_norm": 1.7287795543670654, + "learning_rate": 5.175e-07, + "loss": 0.0162, + "step": 3586 + }, + { + "epoch": 4.2437148772552495, + "grad_norm": 1.155511498451233, + "learning_rate": 5.1625e-07, + "loss": 0.0129, + "step": 3587 + }, + { + "epoch": 4.244897959183674, + "grad_norm": 1.2465813159942627, + "learning_rate": 5.15e-07, + "loss": 0.0152, + "step": 3588 + }, + { + "epoch": 4.246081041112097, + "grad_norm": 1.1818499565124512, + "learning_rate": 5.137500000000001e-07, + "loss": 0.0129, + "step": 3589 + }, + { + "epoch": 4.24726412304052, + "grad_norm": 1.25155770778656, + "learning_rate": 5.125e-07, + "loss": 0.0142, + "step": 3590 + }, + { + "epoch": 4.248447204968944, + "grad_norm": 1.1379051208496094, + "learning_rate": 5.1125e-07, + "loss": 0.0151, + "step": 3591 + }, + { + "epoch": 4.249630286897368, + "grad_norm": 1.462848424911499, + "learning_rate": 5.1e-07, + "loss": 0.0145, + "step": 3592 + }, + { + "epoch": 4.250813368825791, + "grad_norm": 0.7998476028442383, + "learning_rate": 5.0875e-07, + "loss": 0.014, + "step": 3593 + }, + { + "epoch": 4.2519964507542145, + "grad_norm": 1.1074923276901245, + "learning_rate": 5.075000000000001e-07, + "loss": 0.0145, + "step": 3594 + }, + { + "epoch": 4.253179532682639, + "grad_norm": 1.1516653299331665, + "learning_rate": 5.0625e-07, + "loss": 0.0145, + "step": 3595 + }, + { + "epoch": 4.254362614611062, + "grad_norm": 1.1824978590011597, + "learning_rate": 5.05e-07, + "loss": 0.0147, + "step": 3596 + }, + { + "epoch": 4.255545696539485, + "grad_norm": 1.0840556621551514, + "learning_rate": 5.0375e-07, + "loss": 0.017, + "step": 3597 + }, + { + "epoch": 4.256728778467909, + "grad_norm": 1.1449469327926636, + "learning_rate": 5.025000000000001e-07, + "loss": 0.0148, + "step": 3598 + }, + { + "epoch": 4.257911860396333, + "grad_norm": 1.3474303483963013, + "learning_rate": 5.012500000000001e-07, + "loss": 0.0139, + "step": 3599 + }, + { + "epoch": 4.259094942324756, + "grad_norm": 1.1065343618392944, + "learning_rate": 5.000000000000001e-07, + "loss": 0.0109, + "step": 3600 + }, + { + "epoch": 4.2602780242531795, + "grad_norm": 1.1271415948867798, + "learning_rate": 4.987500000000001e-07, + "loss": 0.0142, + "step": 3601 + }, + { + "epoch": 4.261461106181603, + "grad_norm": 1.8976036310195923, + "learning_rate": 4.975000000000001e-07, + "loss": 0.0143, + "step": 3602 + }, + { + "epoch": 4.262644188110027, + "grad_norm": 1.034936785697937, + "learning_rate": 4.9625e-07, + "loss": 0.0124, + "step": 3603 + }, + { + "epoch": 4.26382727003845, + "grad_norm": 1.1512850522994995, + "learning_rate": 4.95e-07, + "loss": 0.0143, + "step": 3604 + }, + { + "epoch": 4.265010351966874, + "grad_norm": 1.1858915090560913, + "learning_rate": 4.9375e-07, + "loss": 0.0141, + "step": 3605 + }, + { + "epoch": 4.266193433895297, + "grad_norm": 1.499306321144104, + "learning_rate": 4.925e-07, + "loss": 0.0167, + "step": 3606 + }, + { + "epoch": 4.267376515823721, + "grad_norm": 1.3264082670211792, + "learning_rate": 4.912500000000001e-07, + "loss": 0.017, + "step": 3607 + }, + { + "epoch": 4.268559597752144, + "grad_norm": 0.9705203175544739, + "learning_rate": 4.900000000000001e-07, + "loss": 0.0141, + "step": 3608 + }, + { + "epoch": 4.269742679680568, + "grad_norm": 1.3086469173431396, + "learning_rate": 4.887500000000001e-07, + "loss": 0.0152, + "step": 3609 + }, + { + "epoch": 4.270925761608991, + "grad_norm": 1.873598337173462, + "learning_rate": 4.875000000000001e-07, + "loss": 0.015, + "step": 3610 + }, + { + "epoch": 4.272108843537415, + "grad_norm": 1.4290244579315186, + "learning_rate": 4.8625e-07, + "loss": 0.0132, + "step": 3611 + }, + { + "epoch": 4.273291925465839, + "grad_norm": 1.3954068422317505, + "learning_rate": 4.85e-07, + "loss": 0.014, + "step": 3612 + }, + { + "epoch": 4.274475007394262, + "grad_norm": 1.4348829984664917, + "learning_rate": 4.8375e-07, + "loss": 0.0158, + "step": 3613 + }, + { + "epoch": 4.275658089322685, + "grad_norm": 1.056287407875061, + "learning_rate": 4.825e-07, + "loss": 0.0118, + "step": 3614 + }, + { + "epoch": 4.276841171251109, + "grad_norm": 1.4585750102996826, + "learning_rate": 4.8125e-07, + "loss": 0.0155, + "step": 3615 + }, + { + "epoch": 4.278024253179533, + "grad_norm": 1.4016259908676147, + "learning_rate": 4.800000000000001e-07, + "loss": 0.0153, + "step": 3616 + }, + { + "epoch": 4.279207335107956, + "grad_norm": 1.4598970413208008, + "learning_rate": 4.787500000000001e-07, + "loss": 0.0142, + "step": 3617 + }, + { + "epoch": 4.280390417036379, + "grad_norm": 1.3074675798416138, + "learning_rate": 4.775000000000001e-07, + "loss": 0.0138, + "step": 3618 + }, + { + "epoch": 4.2815734989648035, + "grad_norm": 0.7384185791015625, + "learning_rate": 4.7625000000000005e-07, + "loss": 0.0138, + "step": 3619 + }, + { + "epoch": 4.282756580893227, + "grad_norm": 1.0707982778549194, + "learning_rate": 4.7500000000000006e-07, + "loss": 0.0146, + "step": 3620 + }, + { + "epoch": 4.28393966282165, + "grad_norm": 1.2352120876312256, + "learning_rate": 4.7375e-07, + "loss": 0.014, + "step": 3621 + }, + { + "epoch": 4.285122744750074, + "grad_norm": 1.3415814638137817, + "learning_rate": 4.7250000000000003e-07, + "loss": 0.0154, + "step": 3622 + }, + { + "epoch": 4.286305826678498, + "grad_norm": 1.2876195907592773, + "learning_rate": 4.7125000000000004e-07, + "loss": 0.019, + "step": 3623 + }, + { + "epoch": 4.287488908606921, + "grad_norm": 1.3201857805252075, + "learning_rate": 4.7000000000000005e-07, + "loss": 0.0124, + "step": 3624 + }, + { + "epoch": 4.288671990535344, + "grad_norm": 1.2237918376922607, + "learning_rate": 4.6875000000000006e-07, + "loss": 0.0148, + "step": 3625 + }, + { + "epoch": 4.2898550724637685, + "grad_norm": 1.879042148590088, + "learning_rate": 4.675e-07, + "loss": 0.0161, + "step": 3626 + }, + { + "epoch": 4.291038154392192, + "grad_norm": 0.8324350118637085, + "learning_rate": 4.6625000000000004e-07, + "loss": 0.0142, + "step": 3627 + }, + { + "epoch": 4.292221236320615, + "grad_norm": 1.1088266372680664, + "learning_rate": 4.6500000000000005e-07, + "loss": 0.0128, + "step": 3628 + }, + { + "epoch": 4.2934043182490385, + "grad_norm": 0.9626941680908203, + "learning_rate": 4.6375000000000006e-07, + "loss": 0.0139, + "step": 3629 + }, + { + "epoch": 4.294587400177463, + "grad_norm": 1.1609612703323364, + "learning_rate": 4.625e-07, + "loss": 0.0121, + "step": 3630 + }, + { + "epoch": 4.295770482105886, + "grad_norm": 1.093837857246399, + "learning_rate": 4.6125000000000003e-07, + "loss": 0.0128, + "step": 3631 + }, + { + "epoch": 4.296953564034309, + "grad_norm": 1.3153479099273682, + "learning_rate": 4.6000000000000004e-07, + "loss": 0.013, + "step": 3632 + }, + { + "epoch": 4.298136645962733, + "grad_norm": 1.2244272232055664, + "learning_rate": 4.5875000000000005e-07, + "loss": 0.0125, + "step": 3633 + }, + { + "epoch": 4.299319727891157, + "grad_norm": 1.2274004220962524, + "learning_rate": 4.575e-07, + "loss": 0.0131, + "step": 3634 + }, + { + "epoch": 4.30050280981958, + "grad_norm": 1.2616418600082397, + "learning_rate": 4.5625e-07, + "loss": 0.013, + "step": 3635 + }, + { + "epoch": 4.301685891748003, + "grad_norm": 1.035027265548706, + "learning_rate": 4.5500000000000004e-07, + "loss": 0.015, + "step": 3636 + }, + { + "epoch": 4.302868973676427, + "grad_norm": 1.2408113479614258, + "learning_rate": 4.5375000000000005e-07, + "loss": 0.0136, + "step": 3637 + }, + { + "epoch": 4.304052055604851, + "grad_norm": 1.1312936544418335, + "learning_rate": 4.525e-07, + "loss": 0.0133, + "step": 3638 + }, + { + "epoch": 4.305235137533274, + "grad_norm": 0.9676356911659241, + "learning_rate": 4.5125e-07, + "loss": 0.0176, + "step": 3639 + }, + { + "epoch": 4.306418219461698, + "grad_norm": 1.3321959972381592, + "learning_rate": 4.5000000000000003e-07, + "loss": 0.011, + "step": 3640 + }, + { + "epoch": 4.307601301390121, + "grad_norm": 1.010885238647461, + "learning_rate": 4.4875000000000004e-07, + "loss": 0.0129, + "step": 3641 + }, + { + "epoch": 4.308784383318545, + "grad_norm": 1.0330592393875122, + "learning_rate": 4.475e-07, + "loss": 0.0136, + "step": 3642 + }, + { + "epoch": 4.309967465246968, + "grad_norm": 1.316184639930725, + "learning_rate": 4.4625e-07, + "loss": 0.0157, + "step": 3643 + }, + { + "epoch": 4.311150547175392, + "grad_norm": 1.2122917175292969, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.0141, + "step": 3644 + }, + { + "epoch": 4.312333629103815, + "grad_norm": 0.851001501083374, + "learning_rate": 4.4375000000000004e-07, + "loss": 0.0134, + "step": 3645 + }, + { + "epoch": 4.313516711032239, + "grad_norm": 1.1388428211212158, + "learning_rate": 4.425e-07, + "loss": 0.0148, + "step": 3646 + }, + { + "epoch": 4.3146997929606625, + "grad_norm": 1.0334173440933228, + "learning_rate": 4.4125e-07, + "loss": 0.0141, + "step": 3647 + }, + { + "epoch": 4.315882874889086, + "grad_norm": 1.0485626459121704, + "learning_rate": 4.4e-07, + "loss": 0.0132, + "step": 3648 + }, + { + "epoch": 4.31706595681751, + "grad_norm": 1.2796655893325806, + "learning_rate": 4.3875000000000003e-07, + "loss": 0.0152, + "step": 3649 + }, + { + "epoch": 4.318249038745933, + "grad_norm": 1.0281015634536743, + "learning_rate": 4.375e-07, + "loss": 0.0128, + "step": 3650 + }, + { + "epoch": 4.319432120674357, + "grad_norm": 1.2790031433105469, + "learning_rate": 4.3625e-07, + "loss": 0.0132, + "step": 3651 + }, + { + "epoch": 4.32061520260278, + "grad_norm": 1.3444329500198364, + "learning_rate": 4.35e-07, + "loss": 0.016, + "step": 3652 + }, + { + "epoch": 4.321798284531204, + "grad_norm": 1.4517184495925903, + "learning_rate": 4.3375000000000003e-07, + "loss": 0.0139, + "step": 3653 + }, + { + "epoch": 4.3229813664596275, + "grad_norm": 1.4574201107025146, + "learning_rate": 4.325e-07, + "loss": 0.0157, + "step": 3654 + }, + { + "epoch": 4.324164448388051, + "grad_norm": 1.375389575958252, + "learning_rate": 4.3125e-07, + "loss": 0.015, + "step": 3655 + }, + { + "epoch": 4.325347530316474, + "grad_norm": 1.4188307523727417, + "learning_rate": 4.3e-07, + "loss": 0.0183, + "step": 3656 + }, + { + "epoch": 4.326530612244898, + "grad_norm": 1.2586756944656372, + "learning_rate": 4.287500000000001e-07, + "loss": 0.0162, + "step": 3657 + }, + { + "epoch": 4.327713694173322, + "grad_norm": 1.3370975255966187, + "learning_rate": 4.275000000000001e-07, + "loss": 0.0141, + "step": 3658 + }, + { + "epoch": 4.328896776101745, + "grad_norm": 1.168445348739624, + "learning_rate": 4.2625000000000005e-07, + "loss": 0.0119, + "step": 3659 + }, + { + "epoch": 4.330079858030168, + "grad_norm": 1.152636170387268, + "learning_rate": 4.2500000000000006e-07, + "loss": 0.017, + "step": 3660 + }, + { + "epoch": 4.3312629399585925, + "grad_norm": 1.043997049331665, + "learning_rate": 4.2375000000000007e-07, + "loss": 0.0166, + "step": 3661 + }, + { + "epoch": 4.332446021887016, + "grad_norm": 0.8921056389808655, + "learning_rate": 4.225000000000001e-07, + "loss": 0.0134, + "step": 3662 + }, + { + "epoch": 4.333629103815439, + "grad_norm": 1.2168604135513306, + "learning_rate": 4.2125000000000004e-07, + "loss": 0.0162, + "step": 3663 + }, + { + "epoch": 4.334812185743862, + "grad_norm": 1.3735309839248657, + "learning_rate": 4.2000000000000006e-07, + "loss": 0.0169, + "step": 3664 + }, + { + "epoch": 4.335995267672287, + "grad_norm": 1.4498330354690552, + "learning_rate": 4.1875000000000007e-07, + "loss": 0.0132, + "step": 3665 + }, + { + "epoch": 4.33717834960071, + "grad_norm": 1.1429939270019531, + "learning_rate": 4.175000000000001e-07, + "loss": 0.0137, + "step": 3666 + }, + { + "epoch": 4.338361431529133, + "grad_norm": 1.3189789056777954, + "learning_rate": 4.1625000000000004e-07, + "loss": 0.013, + "step": 3667 + }, + { + "epoch": 4.3395445134575565, + "grad_norm": 1.0720303058624268, + "learning_rate": 4.1500000000000005e-07, + "loss": 0.0105, + "step": 3668 + }, + { + "epoch": 4.340727595385981, + "grad_norm": 0.8892923593521118, + "learning_rate": 4.1375000000000006e-07, + "loss": 0.0132, + "step": 3669 + }, + { + "epoch": 4.341910677314404, + "grad_norm": 1.2747265100479126, + "learning_rate": 4.125000000000001e-07, + "loss": 0.0197, + "step": 3670 + }, + { + "epoch": 4.343093759242827, + "grad_norm": 1.096947431564331, + "learning_rate": 4.1125000000000003e-07, + "loss": 0.0156, + "step": 3671 + }, + { + "epoch": 4.344276841171251, + "grad_norm": 1.1523938179016113, + "learning_rate": 4.1000000000000004e-07, + "loss": 0.0154, + "step": 3672 + }, + { + "epoch": 4.345459923099675, + "grad_norm": 1.2470190525054932, + "learning_rate": 4.0875000000000006e-07, + "loss": 0.0142, + "step": 3673 + }, + { + "epoch": 4.346643005028098, + "grad_norm": 1.5184391736984253, + "learning_rate": 4.0750000000000007e-07, + "loss": 0.0165, + "step": 3674 + }, + { + "epoch": 4.3478260869565215, + "grad_norm": 1.0692219734191895, + "learning_rate": 4.0625000000000003e-07, + "loss": 0.0139, + "step": 3675 + }, + { + "epoch": 4.349009168884946, + "grad_norm": 1.2974870204925537, + "learning_rate": 4.0500000000000004e-07, + "loss": 0.0139, + "step": 3676 + }, + { + "epoch": 4.350192250813369, + "grad_norm": 1.091141700744629, + "learning_rate": 4.0375000000000005e-07, + "loss": 0.0162, + "step": 3677 + }, + { + "epoch": 4.351375332741792, + "grad_norm": 1.382089376449585, + "learning_rate": 4.0250000000000006e-07, + "loss": 0.0135, + "step": 3678 + }, + { + "epoch": 4.352558414670216, + "grad_norm": 1.0919700860977173, + "learning_rate": 4.0125e-07, + "loss": 0.014, + "step": 3679 + }, + { + "epoch": 4.35374149659864, + "grad_norm": 1.1507911682128906, + "learning_rate": 4.0000000000000003e-07, + "loss": 0.0159, + "step": 3680 + }, + { + "epoch": 4.354924578527063, + "grad_norm": 0.8708051443099976, + "learning_rate": 3.9875000000000005e-07, + "loss": 0.0134, + "step": 3681 + }, + { + "epoch": 4.3561076604554865, + "grad_norm": 1.6522514820098877, + "learning_rate": 3.9750000000000006e-07, + "loss": 0.0152, + "step": 3682 + }, + { + "epoch": 4.35729074238391, + "grad_norm": 1.1387608051300049, + "learning_rate": 3.9625e-07, + "loss": 0.0165, + "step": 3683 + }, + { + "epoch": 4.358473824312334, + "grad_norm": 1.428667664527893, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.0142, + "step": 3684 + }, + { + "epoch": 4.359656906240757, + "grad_norm": 1.2582247257232666, + "learning_rate": 3.9375000000000004e-07, + "loss": 0.0174, + "step": 3685 + }, + { + "epoch": 4.360839988169181, + "grad_norm": 1.189099907875061, + "learning_rate": 3.9250000000000005e-07, + "loss": 0.0146, + "step": 3686 + }, + { + "epoch": 4.362023070097604, + "grad_norm": 1.3747448921203613, + "learning_rate": 3.9125e-07, + "loss": 0.0192, + "step": 3687 + }, + { + "epoch": 4.363206152026028, + "grad_norm": 1.2179200649261475, + "learning_rate": 3.9e-07, + "loss": 0.0129, + "step": 3688 + }, + { + "epoch": 4.364389233954451, + "grad_norm": 1.786212682723999, + "learning_rate": 3.8875000000000004e-07, + "loss": 0.0145, + "step": 3689 + }, + { + "epoch": 4.365572315882875, + "grad_norm": 1.1598351001739502, + "learning_rate": 3.8750000000000005e-07, + "loss": 0.0135, + "step": 3690 + }, + { + "epoch": 4.366755397811298, + "grad_norm": 1.675309181213379, + "learning_rate": 3.8625e-07, + "loss": 0.0121, + "step": 3691 + }, + { + "epoch": 4.367938479739722, + "grad_norm": 1.5713818073272705, + "learning_rate": 3.85e-07, + "loss": 0.0144, + "step": 3692 + }, + { + "epoch": 4.369121561668146, + "grad_norm": 0.9869341850280762, + "learning_rate": 3.8375000000000003e-07, + "loss": 0.011, + "step": 3693 + }, + { + "epoch": 4.370304643596569, + "grad_norm": 1.1807652711868286, + "learning_rate": 3.8250000000000004e-07, + "loss": 0.0152, + "step": 3694 + }, + { + "epoch": 4.371487725524992, + "grad_norm": 0.8826397061347961, + "learning_rate": 3.8125e-07, + "loss": 0.0125, + "step": 3695 + }, + { + "epoch": 4.372670807453416, + "grad_norm": 1.1989405155181885, + "learning_rate": 3.8e-07, + "loss": 0.0145, + "step": 3696 + }, + { + "epoch": 4.37385388938184, + "grad_norm": 1.662712574005127, + "learning_rate": 3.7875000000000003e-07, + "loss": 0.0151, + "step": 3697 + }, + { + "epoch": 4.375036971310263, + "grad_norm": 1.3821622133255005, + "learning_rate": 3.7750000000000004e-07, + "loss": 0.0151, + "step": 3698 + }, + { + "epoch": 4.376220053238686, + "grad_norm": 1.1394133567810059, + "learning_rate": 3.7625e-07, + "loss": 0.0111, + "step": 3699 + }, + { + "epoch": 4.3774031351671105, + "grad_norm": 1.1627509593963623, + "learning_rate": 3.75e-07, + "loss": 0.0123, + "step": 3700 + }, + { + "epoch": 4.378586217095534, + "grad_norm": 1.2807142734527588, + "learning_rate": 3.7375e-07, + "loss": 0.0149, + "step": 3701 + }, + { + "epoch": 4.379769299023957, + "grad_norm": 0.8952203989028931, + "learning_rate": 3.7250000000000003e-07, + "loss": 0.0147, + "step": 3702 + }, + { + "epoch": 4.380952380952381, + "grad_norm": 1.2361639738082886, + "learning_rate": 3.7125e-07, + "loss": 0.0153, + "step": 3703 + }, + { + "epoch": 4.382135462880805, + "grad_norm": 1.5076000690460205, + "learning_rate": 3.7e-07, + "loss": 0.0135, + "step": 3704 + }, + { + "epoch": 4.383318544809228, + "grad_norm": 1.0191949605941772, + "learning_rate": 3.6875e-07, + "loss": 0.0118, + "step": 3705 + }, + { + "epoch": 4.384501626737651, + "grad_norm": 1.2322627305984497, + "learning_rate": 3.6750000000000003e-07, + "loss": 0.0142, + "step": 3706 + }, + { + "epoch": 4.3856847086660755, + "grad_norm": 1.7396570444107056, + "learning_rate": 3.6625e-07, + "loss": 0.0142, + "step": 3707 + }, + { + "epoch": 4.386867790594499, + "grad_norm": 1.6677711009979248, + "learning_rate": 3.65e-07, + "loss": 0.0136, + "step": 3708 + }, + { + "epoch": 4.388050872522922, + "grad_norm": 1.2260419130325317, + "learning_rate": 3.6375e-07, + "loss": 0.0165, + "step": 3709 + }, + { + "epoch": 4.3892339544513455, + "grad_norm": 1.1949443817138672, + "learning_rate": 3.625e-07, + "loss": 0.0135, + "step": 3710 + }, + { + "epoch": 4.39041703637977, + "grad_norm": 1.1633620262145996, + "learning_rate": 3.6125e-07, + "loss": 0.0127, + "step": 3711 + }, + { + "epoch": 4.391600118308193, + "grad_norm": 1.4265475273132324, + "learning_rate": 3.6e-07, + "loss": 0.0163, + "step": 3712 + }, + { + "epoch": 4.392783200236616, + "grad_norm": 1.2298752069473267, + "learning_rate": 3.5875e-07, + "loss": 0.0139, + "step": 3713 + }, + { + "epoch": 4.39396628216504, + "grad_norm": 1.3417061567306519, + "learning_rate": 3.575e-07, + "loss": 0.0157, + "step": 3714 + }, + { + "epoch": 4.395149364093464, + "grad_norm": 1.2389531135559082, + "learning_rate": 3.5625e-07, + "loss": 0.0126, + "step": 3715 + }, + { + "epoch": 4.396332446021887, + "grad_norm": 1.300001859664917, + "learning_rate": 3.55e-07, + "loss": 0.0156, + "step": 3716 + }, + { + "epoch": 4.39751552795031, + "grad_norm": 1.7401446104049683, + "learning_rate": 3.5375e-07, + "loss": 0.0174, + "step": 3717 + }, + { + "epoch": 4.398698609878734, + "grad_norm": 1.0672249794006348, + "learning_rate": 3.525e-07, + "loss": 0.011, + "step": 3718 + }, + { + "epoch": 4.399881691807158, + "grad_norm": 1.4920117855072021, + "learning_rate": 3.512500000000001e-07, + "loss": 0.0116, + "step": 3719 + }, + { + "epoch": 4.401064773735581, + "grad_norm": 1.0158312320709229, + "learning_rate": 3.5000000000000004e-07, + "loss": 0.0125, + "step": 3720 + }, + { + "epoch": 4.402247855664005, + "grad_norm": 1.1158419847488403, + "learning_rate": 3.4875000000000005e-07, + "loss": 0.0145, + "step": 3721 + }, + { + "epoch": 4.403430937592428, + "grad_norm": 1.0790088176727295, + "learning_rate": 3.4750000000000006e-07, + "loss": 0.0133, + "step": 3722 + }, + { + "epoch": 4.404614019520852, + "grad_norm": 1.1396714448928833, + "learning_rate": 3.4625000000000007e-07, + "loss": 0.0175, + "step": 3723 + }, + { + "epoch": 4.405797101449275, + "grad_norm": 1.4706292152404785, + "learning_rate": 3.4500000000000003e-07, + "loss": 0.0132, + "step": 3724 + }, + { + "epoch": 4.406980183377699, + "grad_norm": 1.1139123439788818, + "learning_rate": 3.4375000000000004e-07, + "loss": 0.0124, + "step": 3725 + }, + { + "epoch": 4.408163265306122, + "grad_norm": 1.1300973892211914, + "learning_rate": 3.4250000000000006e-07, + "loss": 0.0115, + "step": 3726 + }, + { + "epoch": 4.409346347234546, + "grad_norm": 1.1652758121490479, + "learning_rate": 3.4125000000000007e-07, + "loss": 0.0172, + "step": 3727 + }, + { + "epoch": 4.4105294291629695, + "grad_norm": 0.8881568908691406, + "learning_rate": 3.4000000000000003e-07, + "loss": 0.0138, + "step": 3728 + }, + { + "epoch": 4.411712511091393, + "grad_norm": 1.4512152671813965, + "learning_rate": 3.3875000000000004e-07, + "loss": 0.0133, + "step": 3729 + }, + { + "epoch": 4.412895593019817, + "grad_norm": 1.132643222808838, + "learning_rate": 3.3750000000000005e-07, + "loss": 0.0128, + "step": 3730 + }, + { + "epoch": 4.41407867494824, + "grad_norm": 0.9694779515266418, + "learning_rate": 3.3625000000000006e-07, + "loss": 0.0129, + "step": 3731 + }, + { + "epoch": 4.415261756876664, + "grad_norm": 1.1862062215805054, + "learning_rate": 3.35e-07, + "loss": 0.0144, + "step": 3732 + }, + { + "epoch": 4.416444838805087, + "grad_norm": 1.4223309755325317, + "learning_rate": 3.3375000000000003e-07, + "loss": 0.0139, + "step": 3733 + }, + { + "epoch": 4.417627920733511, + "grad_norm": 1.1662510633468628, + "learning_rate": 3.3250000000000005e-07, + "loss": 0.0147, + "step": 3734 + }, + { + "epoch": 4.4188110026619345, + "grad_norm": 1.1835424900054932, + "learning_rate": 3.3125000000000006e-07, + "loss": 0.0149, + "step": 3735 + }, + { + "epoch": 4.419994084590358, + "grad_norm": 0.7946937084197998, + "learning_rate": 3.3e-07, + "loss": 0.0095, + "step": 3736 + }, + { + "epoch": 4.421177166518781, + "grad_norm": 1.062906265258789, + "learning_rate": 3.2875000000000003e-07, + "loss": 0.0142, + "step": 3737 + }, + { + "epoch": 4.422360248447205, + "grad_norm": 1.3007041215896606, + "learning_rate": 3.2750000000000004e-07, + "loss": 0.0146, + "step": 3738 + }, + { + "epoch": 4.423543330375629, + "grad_norm": 1.307305932044983, + "learning_rate": 3.2625000000000005e-07, + "loss": 0.0128, + "step": 3739 + }, + { + "epoch": 4.424726412304052, + "grad_norm": 1.5250734090805054, + "learning_rate": 3.25e-07, + "loss": 0.0158, + "step": 3740 + }, + { + "epoch": 4.425909494232475, + "grad_norm": 1.8355926275253296, + "learning_rate": 3.2375e-07, + "loss": 0.0131, + "step": 3741 + }, + { + "epoch": 4.4270925761608995, + "grad_norm": 1.2511670589447021, + "learning_rate": 3.2250000000000004e-07, + "loss": 0.0142, + "step": 3742 + }, + { + "epoch": 4.428275658089323, + "grad_norm": 1.0021952390670776, + "learning_rate": 3.2125000000000005e-07, + "loss": 0.0117, + "step": 3743 + }, + { + "epoch": 4.429458740017746, + "grad_norm": 1.2392067909240723, + "learning_rate": 3.2e-07, + "loss": 0.0131, + "step": 3744 + }, + { + "epoch": 4.430641821946169, + "grad_norm": 1.1209744215011597, + "learning_rate": 3.1875e-07, + "loss": 0.0118, + "step": 3745 + }, + { + "epoch": 4.431824903874594, + "grad_norm": 0.8967317938804626, + "learning_rate": 3.1750000000000003e-07, + "loss": 0.0117, + "step": 3746 + }, + { + "epoch": 4.433007985803017, + "grad_norm": 1.958120584487915, + "learning_rate": 3.1625000000000004e-07, + "loss": 0.0131, + "step": 3747 + }, + { + "epoch": 4.43419106773144, + "grad_norm": 1.0072057247161865, + "learning_rate": 3.15e-07, + "loss": 0.0103, + "step": 3748 + }, + { + "epoch": 4.4353741496598635, + "grad_norm": 0.9765480756759644, + "learning_rate": 3.1375e-07, + "loss": 0.0151, + "step": 3749 + }, + { + "epoch": 4.436557231588288, + "grad_norm": 0.964718759059906, + "learning_rate": 3.125e-07, + "loss": 0.0117, + "step": 3750 + }, + { + "epoch": 4.437740313516711, + "grad_norm": 1.114830732345581, + "learning_rate": 3.1125000000000004e-07, + "loss": 0.0143, + "step": 3751 + }, + { + "epoch": 4.438923395445134, + "grad_norm": 0.8220805525779724, + "learning_rate": 3.1000000000000005e-07, + "loss": 0.0113, + "step": 3752 + }, + { + "epoch": 4.440106477373558, + "grad_norm": 0.8409322500228882, + "learning_rate": 3.0875e-07, + "loss": 0.0147, + "step": 3753 + }, + { + "epoch": 4.441289559301982, + "grad_norm": 1.1091654300689697, + "learning_rate": 3.075e-07, + "loss": 0.0176, + "step": 3754 + }, + { + "epoch": 4.442472641230405, + "grad_norm": 1.1023364067077637, + "learning_rate": 3.0625000000000003e-07, + "loss": 0.0151, + "step": 3755 + }, + { + "epoch": 4.4436557231588285, + "grad_norm": 0.9273171424865723, + "learning_rate": 3.0500000000000004e-07, + "loss": 0.0139, + "step": 3756 + }, + { + "epoch": 4.444838805087253, + "grad_norm": 0.9335495829582214, + "learning_rate": 3.0375e-07, + "loss": 0.0118, + "step": 3757 + }, + { + "epoch": 4.446021887015676, + "grad_norm": 1.4977680444717407, + "learning_rate": 3.025e-07, + "loss": 0.0145, + "step": 3758 + }, + { + "epoch": 4.447204968944099, + "grad_norm": 1.3055846691131592, + "learning_rate": 3.0125000000000003e-07, + "loss": 0.0129, + "step": 3759 + }, + { + "epoch": 4.448388050872523, + "grad_norm": 1.1620619297027588, + "learning_rate": 3.0000000000000004e-07, + "loss": 0.0147, + "step": 3760 + }, + { + "epoch": 4.449571132800947, + "grad_norm": 1.1118556261062622, + "learning_rate": 2.9875e-07, + "loss": 0.0146, + "step": 3761 + }, + { + "epoch": 4.45075421472937, + "grad_norm": 1.4309312105178833, + "learning_rate": 2.975e-07, + "loss": 0.0178, + "step": 3762 + }, + { + "epoch": 4.4519372966577935, + "grad_norm": 1.5483176708221436, + "learning_rate": 2.9625e-07, + "loss": 0.0139, + "step": 3763 + }, + { + "epoch": 4.453120378586217, + "grad_norm": 1.298574447631836, + "learning_rate": 2.9500000000000003e-07, + "loss": 0.0154, + "step": 3764 + }, + { + "epoch": 4.454303460514641, + "grad_norm": 1.3794598579406738, + "learning_rate": 2.9375e-07, + "loss": 0.0164, + "step": 3765 + }, + { + "epoch": 4.455486542443064, + "grad_norm": 0.9628676176071167, + "learning_rate": 2.9250000000000006e-07, + "loss": 0.0121, + "step": 3766 + }, + { + "epoch": 4.456669624371488, + "grad_norm": 1.1678799390792847, + "learning_rate": 2.9125e-07, + "loss": 0.0114, + "step": 3767 + }, + { + "epoch": 4.457852706299911, + "grad_norm": 1.1425331830978394, + "learning_rate": 2.9000000000000003e-07, + "loss": 0.0129, + "step": 3768 + }, + { + "epoch": 4.459035788228335, + "grad_norm": 0.9871340394020081, + "learning_rate": 2.8875000000000004e-07, + "loss": 0.0102, + "step": 3769 + }, + { + "epoch": 4.460218870156758, + "grad_norm": 1.0501760244369507, + "learning_rate": 2.8750000000000005e-07, + "loss": 0.0124, + "step": 3770 + }, + { + "epoch": 4.461401952085182, + "grad_norm": 1.1085304021835327, + "learning_rate": 2.8625e-07, + "loss": 0.0147, + "step": 3771 + }, + { + "epoch": 4.462585034013605, + "grad_norm": 1.0973775386810303, + "learning_rate": 2.85e-07, + "loss": 0.0141, + "step": 3772 + }, + { + "epoch": 4.463768115942029, + "grad_norm": 0.9608280062675476, + "learning_rate": 2.8375000000000004e-07, + "loss": 0.0127, + "step": 3773 + }, + { + "epoch": 4.464951197870453, + "grad_norm": 1.6411699056625366, + "learning_rate": 2.8250000000000005e-07, + "loss": 0.014, + "step": 3774 + }, + { + "epoch": 4.466134279798876, + "grad_norm": 1.0823729038238525, + "learning_rate": 2.8125e-07, + "loss": 0.0125, + "step": 3775 + }, + { + "epoch": 4.467317361727299, + "grad_norm": 1.3735905885696411, + "learning_rate": 2.8e-07, + "loss": 0.0135, + "step": 3776 + }, + { + "epoch": 4.468500443655723, + "grad_norm": 1.179389476776123, + "learning_rate": 2.7875000000000003e-07, + "loss": 0.0167, + "step": 3777 + }, + { + "epoch": 4.469683525584147, + "grad_norm": 1.2777912616729736, + "learning_rate": 2.7750000000000004e-07, + "loss": 0.0144, + "step": 3778 + }, + { + "epoch": 4.47086660751257, + "grad_norm": 1.3312983512878418, + "learning_rate": 2.7625e-07, + "loss": 0.0144, + "step": 3779 + }, + { + "epoch": 4.472049689440993, + "grad_norm": 1.6353635787963867, + "learning_rate": 2.75e-07, + "loss": 0.0123, + "step": 3780 + }, + { + "epoch": 4.4732327713694175, + "grad_norm": 1.13539457321167, + "learning_rate": 2.7375e-07, + "loss": 0.0123, + "step": 3781 + }, + { + "epoch": 4.474415853297841, + "grad_norm": 2.0901618003845215, + "learning_rate": 2.7250000000000004e-07, + "loss": 0.0119, + "step": 3782 + }, + { + "epoch": 4.475598935226264, + "grad_norm": 1.0477805137634277, + "learning_rate": 2.7125e-07, + "loss": 0.0149, + "step": 3783 + }, + { + "epoch": 4.476782017154688, + "grad_norm": 0.9666970372200012, + "learning_rate": 2.7e-07, + "loss": 0.0111, + "step": 3784 + }, + { + "epoch": 4.477965099083112, + "grad_norm": 0.831697404384613, + "learning_rate": 2.6875e-07, + "loss": 0.0149, + "step": 3785 + }, + { + "epoch": 4.479148181011535, + "grad_norm": 0.9649348258972168, + "learning_rate": 2.6750000000000003e-07, + "loss": 0.0133, + "step": 3786 + }, + { + "epoch": 4.480331262939958, + "grad_norm": 1.4897058010101318, + "learning_rate": 2.6625e-07, + "loss": 0.014, + "step": 3787 + }, + { + "epoch": 4.4815143448683825, + "grad_norm": 1.5082415342330933, + "learning_rate": 2.65e-07, + "loss": 0.0125, + "step": 3788 + }, + { + "epoch": 4.482697426796806, + "grad_norm": 0.90255206823349, + "learning_rate": 2.6375e-07, + "loss": 0.0131, + "step": 3789 + }, + { + "epoch": 4.483880508725229, + "grad_norm": 1.1254191398620605, + "learning_rate": 2.6250000000000003e-07, + "loss": 0.0115, + "step": 3790 + }, + { + "epoch": 4.4850635906536525, + "grad_norm": 1.320709466934204, + "learning_rate": 2.6125e-07, + "loss": 0.0129, + "step": 3791 + }, + { + "epoch": 4.486246672582077, + "grad_norm": 1.0462384223937988, + "learning_rate": 2.6e-07, + "loss": 0.011, + "step": 3792 + }, + { + "epoch": 4.4874297545105, + "grad_norm": 1.5363869667053223, + "learning_rate": 2.5875e-07, + "loss": 0.0135, + "step": 3793 + }, + { + "epoch": 4.488612836438923, + "grad_norm": 1.739203691482544, + "learning_rate": 2.575e-07, + "loss": 0.02, + "step": 3794 + }, + { + "epoch": 4.489795918367347, + "grad_norm": 1.1987860202789307, + "learning_rate": 2.5625e-07, + "loss": 0.0141, + "step": 3795 + }, + { + "epoch": 4.490979000295771, + "grad_norm": 1.308578610420227, + "learning_rate": 2.55e-07, + "loss": 0.0157, + "step": 3796 + }, + { + "epoch": 4.492162082224194, + "grad_norm": 1.013198971748352, + "learning_rate": 2.5375000000000006e-07, + "loss": 0.0144, + "step": 3797 + }, + { + "epoch": 4.493345164152617, + "grad_norm": 1.2182115316390991, + "learning_rate": 2.525e-07, + "loss": 0.0126, + "step": 3798 + }, + { + "epoch": 4.494528246081041, + "grad_norm": 1.7988308668136597, + "learning_rate": 2.5125000000000003e-07, + "loss": 0.014, + "step": 3799 + }, + { + "epoch": 4.495711328009465, + "grad_norm": 1.4154717922210693, + "learning_rate": 2.5000000000000004e-07, + "loss": 0.0142, + "step": 3800 + }, + { + "epoch": 4.496894409937888, + "grad_norm": 0.9756066203117371, + "learning_rate": 2.4875000000000005e-07, + "loss": 0.0122, + "step": 3801 + }, + { + "epoch": 4.498077491866312, + "grad_norm": 1.5958011150360107, + "learning_rate": 2.475e-07, + "loss": 0.0146, + "step": 3802 + }, + { + "epoch": 4.499260573794735, + "grad_norm": 1.4842623472213745, + "learning_rate": 2.4625e-07, + "loss": 0.0177, + "step": 3803 + }, + { + "epoch": 4.500443655723159, + "grad_norm": 1.0267225503921509, + "learning_rate": 2.4500000000000004e-07, + "loss": 0.0125, + "step": 3804 + }, + { + "epoch": 4.501626737651582, + "grad_norm": 1.430670142173767, + "learning_rate": 2.4375000000000005e-07, + "loss": 0.0161, + "step": 3805 + }, + { + "epoch": 4.502809819580006, + "grad_norm": 1.3319512605667114, + "learning_rate": 2.425e-07, + "loss": 0.0143, + "step": 3806 + }, + { + "epoch": 4.503992901508429, + "grad_norm": 1.1301220655441284, + "learning_rate": 2.4125e-07, + "loss": 0.0142, + "step": 3807 + }, + { + "epoch": 4.505175983436853, + "grad_norm": 1.3003913164138794, + "learning_rate": 2.4000000000000003e-07, + "loss": 0.0116, + "step": 3808 + }, + { + "epoch": 4.5063590653652765, + "grad_norm": 1.5558748245239258, + "learning_rate": 2.3875000000000004e-07, + "loss": 0.0171, + "step": 3809 + }, + { + "epoch": 4.5075421472937, + "grad_norm": 1.08849036693573, + "learning_rate": 2.3750000000000003e-07, + "loss": 0.0125, + "step": 3810 + }, + { + "epoch": 4.508725229222124, + "grad_norm": 1.1083614826202393, + "learning_rate": 2.3625000000000001e-07, + "loss": 0.0143, + "step": 3811 + }, + { + "epoch": 4.509908311150547, + "grad_norm": 1.076951265335083, + "learning_rate": 2.3500000000000003e-07, + "loss": 0.0144, + "step": 3812 + }, + { + "epoch": 4.511091393078971, + "grad_norm": 0.7894152402877808, + "learning_rate": 2.3375e-07, + "loss": 0.0168, + "step": 3813 + }, + { + "epoch": 4.512274475007394, + "grad_norm": 1.0263798236846924, + "learning_rate": 2.3250000000000002e-07, + "loss": 0.0127, + "step": 3814 + }, + { + "epoch": 4.513457556935818, + "grad_norm": 1.0314431190490723, + "learning_rate": 2.3125e-07, + "loss": 0.0125, + "step": 3815 + }, + { + "epoch": 4.5146406388642415, + "grad_norm": 1.2768052816390991, + "learning_rate": 2.3000000000000002e-07, + "loss": 0.0131, + "step": 3816 + }, + { + "epoch": 4.515823720792665, + "grad_norm": 1.0892399549484253, + "learning_rate": 2.2875e-07, + "loss": 0.0129, + "step": 3817 + }, + { + "epoch": 4.517006802721088, + "grad_norm": 1.1472368240356445, + "learning_rate": 2.2750000000000002e-07, + "loss": 0.0153, + "step": 3818 + }, + { + "epoch": 4.518189884649512, + "grad_norm": 0.8188145160675049, + "learning_rate": 2.2625e-07, + "loss": 0.0114, + "step": 3819 + }, + { + "epoch": 4.519372966577936, + "grad_norm": 1.4394800662994385, + "learning_rate": 2.2500000000000002e-07, + "loss": 0.0126, + "step": 3820 + }, + { + "epoch": 4.520556048506359, + "grad_norm": 1.325788974761963, + "learning_rate": 2.2375e-07, + "loss": 0.0155, + "step": 3821 + }, + { + "epoch": 4.521739130434782, + "grad_norm": 0.7463934421539307, + "learning_rate": 2.2250000000000001e-07, + "loss": 0.0134, + "step": 3822 + }, + { + "epoch": 4.5229222123632065, + "grad_norm": 1.2386423349380493, + "learning_rate": 2.2125e-07, + "loss": 0.0182, + "step": 3823 + }, + { + "epoch": 4.52410529429163, + "grad_norm": 1.0408538579940796, + "learning_rate": 2.2e-07, + "loss": 0.0137, + "step": 3824 + }, + { + "epoch": 4.525288376220053, + "grad_norm": 0.7979284524917603, + "learning_rate": 2.1875e-07, + "loss": 0.0108, + "step": 3825 + }, + { + "epoch": 4.526471458148476, + "grad_norm": 1.2900367975234985, + "learning_rate": 2.175e-07, + "loss": 0.0128, + "step": 3826 + }, + { + "epoch": 4.527654540076901, + "grad_norm": 1.1793699264526367, + "learning_rate": 2.1625e-07, + "loss": 0.0135, + "step": 3827 + }, + { + "epoch": 4.528837622005324, + "grad_norm": 1.3840960264205933, + "learning_rate": 2.15e-07, + "loss": 0.0105, + "step": 3828 + }, + { + "epoch": 4.530020703933747, + "grad_norm": 1.3441182374954224, + "learning_rate": 2.1375000000000004e-07, + "loss": 0.0131, + "step": 3829 + }, + { + "epoch": 4.531203785862171, + "grad_norm": 1.0488234758377075, + "learning_rate": 2.1250000000000003e-07, + "loss": 0.0174, + "step": 3830 + }, + { + "epoch": 4.532386867790595, + "grad_norm": 1.7993664741516113, + "learning_rate": 2.1125000000000004e-07, + "loss": 0.0153, + "step": 3831 + }, + { + "epoch": 4.533569949719018, + "grad_norm": 0.912959635257721, + "learning_rate": 2.1000000000000003e-07, + "loss": 0.0132, + "step": 3832 + }, + { + "epoch": 4.534753031647441, + "grad_norm": 1.4288142919540405, + "learning_rate": 2.0875000000000004e-07, + "loss": 0.0129, + "step": 3833 + }, + { + "epoch": 4.535936113575865, + "grad_norm": 1.3132165670394897, + "learning_rate": 2.0750000000000003e-07, + "loss": 0.0118, + "step": 3834 + }, + { + "epoch": 4.537119195504289, + "grad_norm": 1.161919116973877, + "learning_rate": 2.0625000000000004e-07, + "loss": 0.0126, + "step": 3835 + }, + { + "epoch": 4.538302277432712, + "grad_norm": 1.0462740659713745, + "learning_rate": 2.0500000000000002e-07, + "loss": 0.0136, + "step": 3836 + }, + { + "epoch": 4.5394853593611355, + "grad_norm": 0.8959895372390747, + "learning_rate": 2.0375000000000003e-07, + "loss": 0.0113, + "step": 3837 + }, + { + "epoch": 4.54066844128956, + "grad_norm": 0.9805852770805359, + "learning_rate": 2.0250000000000002e-07, + "loss": 0.0141, + "step": 3838 + }, + { + "epoch": 4.541851523217983, + "grad_norm": 0.9318886995315552, + "learning_rate": 2.0125000000000003e-07, + "loss": 0.0121, + "step": 3839 + }, + { + "epoch": 4.543034605146406, + "grad_norm": 1.0862581729888916, + "learning_rate": 2.0000000000000002e-07, + "loss": 0.0109, + "step": 3840 + }, + { + "epoch": 4.54421768707483, + "grad_norm": 1.12823486328125, + "learning_rate": 1.9875000000000003e-07, + "loss": 0.0143, + "step": 3841 + }, + { + "epoch": 4.545400769003254, + "grad_norm": 0.8125132322311401, + "learning_rate": 1.9750000000000001e-07, + "loss": 0.0114, + "step": 3842 + }, + { + "epoch": 4.546583850931677, + "grad_norm": 1.1634286642074585, + "learning_rate": 1.9625000000000003e-07, + "loss": 0.0109, + "step": 3843 + }, + { + "epoch": 4.5477669328601005, + "grad_norm": 1.3765102624893188, + "learning_rate": 1.95e-07, + "loss": 0.0132, + "step": 3844 + }, + { + "epoch": 4.548950014788524, + "grad_norm": 1.0627636909484863, + "learning_rate": 1.9375000000000002e-07, + "loss": 0.0128, + "step": 3845 + }, + { + "epoch": 4.550133096716948, + "grad_norm": 1.3265103101730347, + "learning_rate": 1.925e-07, + "loss": 0.0128, + "step": 3846 + }, + { + "epoch": 4.551316178645371, + "grad_norm": 1.1801010370254517, + "learning_rate": 1.9125000000000002e-07, + "loss": 0.0118, + "step": 3847 + }, + { + "epoch": 4.552499260573795, + "grad_norm": 1.7943882942199707, + "learning_rate": 1.9e-07, + "loss": 0.0145, + "step": 3848 + }, + { + "epoch": 4.553682342502218, + "grad_norm": 1.179008960723877, + "learning_rate": 1.8875000000000002e-07, + "loss": 0.0155, + "step": 3849 + }, + { + "epoch": 4.554865424430642, + "grad_norm": 1.3862535953521729, + "learning_rate": 1.875e-07, + "loss": 0.013, + "step": 3850 + }, + { + "epoch": 4.556048506359065, + "grad_norm": 1.1757287979125977, + "learning_rate": 1.8625000000000002e-07, + "loss": 0.0127, + "step": 3851 + }, + { + "epoch": 4.557231588287489, + "grad_norm": 1.0114079713821411, + "learning_rate": 1.85e-07, + "loss": 0.0119, + "step": 3852 + }, + { + "epoch": 4.558414670215912, + "grad_norm": 1.2452207803726196, + "learning_rate": 1.8375000000000001e-07, + "loss": 0.0123, + "step": 3853 + }, + { + "epoch": 4.559597752144336, + "grad_norm": 1.5236400365829468, + "learning_rate": 1.825e-07, + "loss": 0.0132, + "step": 3854 + }, + { + "epoch": 4.56078083407276, + "grad_norm": 1.4297497272491455, + "learning_rate": 1.8125e-07, + "loss": 0.0139, + "step": 3855 + }, + { + "epoch": 4.561963916001183, + "grad_norm": 1.083322286605835, + "learning_rate": 1.8e-07, + "loss": 0.0123, + "step": 3856 + }, + { + "epoch": 4.563146997929607, + "grad_norm": 1.768494725227356, + "learning_rate": 1.7875e-07, + "loss": 0.0125, + "step": 3857 + }, + { + "epoch": 4.56433007985803, + "grad_norm": 1.0100175142288208, + "learning_rate": 1.775e-07, + "loss": 0.0123, + "step": 3858 + }, + { + "epoch": 4.565513161786454, + "grad_norm": 1.1412909030914307, + "learning_rate": 1.7625e-07, + "loss": 0.0141, + "step": 3859 + }, + { + "epoch": 4.566696243714877, + "grad_norm": 1.3609035015106201, + "learning_rate": 1.7500000000000002e-07, + "loss": 0.0154, + "step": 3860 + }, + { + "epoch": 4.5678793256433, + "grad_norm": 1.4517232179641724, + "learning_rate": 1.7375000000000003e-07, + "loss": 0.0171, + "step": 3861 + }, + { + "epoch": 4.5690624075717245, + "grad_norm": 1.0010218620300293, + "learning_rate": 1.7250000000000002e-07, + "loss": 0.0135, + "step": 3862 + }, + { + "epoch": 4.570245489500148, + "grad_norm": 0.8790861964225769, + "learning_rate": 1.7125000000000003e-07, + "loss": 0.0126, + "step": 3863 + }, + { + "epoch": 4.571428571428571, + "grad_norm": 1.2164216041564941, + "learning_rate": 1.7000000000000001e-07, + "loss": 0.0159, + "step": 3864 + }, + { + "epoch": 4.572611653356995, + "grad_norm": 1.1402498483657837, + "learning_rate": 1.6875000000000003e-07, + "loss": 0.0125, + "step": 3865 + }, + { + "epoch": 4.573794735285419, + "grad_norm": 1.5375196933746338, + "learning_rate": 1.675e-07, + "loss": 0.0119, + "step": 3866 + }, + { + "epoch": 4.574977817213842, + "grad_norm": 0.9963312149047852, + "learning_rate": 1.6625000000000002e-07, + "loss": 0.0121, + "step": 3867 + }, + { + "epoch": 4.576160899142265, + "grad_norm": 1.1417778730392456, + "learning_rate": 1.65e-07, + "loss": 0.0129, + "step": 3868 + }, + { + "epoch": 4.5773439810706895, + "grad_norm": 1.4033547639846802, + "learning_rate": 1.6375000000000002e-07, + "loss": 0.0124, + "step": 3869 + }, + { + "epoch": 4.578527062999113, + "grad_norm": 1.1692861318588257, + "learning_rate": 1.625e-07, + "loss": 0.0125, + "step": 3870 + }, + { + "epoch": 4.579710144927536, + "grad_norm": 1.110651969909668, + "learning_rate": 1.6125000000000002e-07, + "loss": 0.0133, + "step": 3871 + }, + { + "epoch": 4.5808932268559595, + "grad_norm": 1.123038649559021, + "learning_rate": 1.6e-07, + "loss": 0.0137, + "step": 3872 + }, + { + "epoch": 4.582076308784384, + "grad_norm": 1.338476538658142, + "learning_rate": 1.5875000000000002e-07, + "loss": 0.0128, + "step": 3873 + }, + { + "epoch": 4.583259390712807, + "grad_norm": 1.1311776638031006, + "learning_rate": 1.575e-07, + "loss": 0.0133, + "step": 3874 + }, + { + "epoch": 4.58444247264123, + "grad_norm": 1.2533353567123413, + "learning_rate": 1.5625e-07, + "loss": 0.0172, + "step": 3875 + }, + { + "epoch": 4.585625554569654, + "grad_norm": 1.0440913438796997, + "learning_rate": 1.5500000000000002e-07, + "loss": 0.0135, + "step": 3876 + }, + { + "epoch": 4.586808636498078, + "grad_norm": 1.2165313959121704, + "learning_rate": 1.5375e-07, + "loss": 0.0168, + "step": 3877 + }, + { + "epoch": 4.587991718426501, + "grad_norm": 1.0756720304489136, + "learning_rate": 1.5250000000000002e-07, + "loss": 0.013, + "step": 3878 + }, + { + "epoch": 4.589174800354924, + "grad_norm": 1.4220173358917236, + "learning_rate": 1.5125e-07, + "loss": 0.0162, + "step": 3879 + }, + { + "epoch": 4.590357882283348, + "grad_norm": 1.3133206367492676, + "learning_rate": 1.5000000000000002e-07, + "loss": 0.0137, + "step": 3880 + }, + { + "epoch": 4.591540964211772, + "grad_norm": 2.1241555213928223, + "learning_rate": 1.4875e-07, + "loss": 0.0178, + "step": 3881 + }, + { + "epoch": 4.592724046140195, + "grad_norm": 1.3150873184204102, + "learning_rate": 1.4750000000000002e-07, + "loss": 0.0156, + "step": 3882 + }, + { + "epoch": 4.593907128068619, + "grad_norm": 1.2382774353027344, + "learning_rate": 1.4625000000000003e-07, + "loss": 0.0136, + "step": 3883 + }, + { + "epoch": 4.595090209997043, + "grad_norm": 1.7870086431503296, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.0126, + "step": 3884 + }, + { + "epoch": 4.596273291925466, + "grad_norm": 0.9954041242599487, + "learning_rate": 1.4375000000000003e-07, + "loss": 0.0139, + "step": 3885 + }, + { + "epoch": 4.597456373853889, + "grad_norm": 1.1863747835159302, + "learning_rate": 1.425e-07, + "loss": 0.0147, + "step": 3886 + }, + { + "epoch": 4.598639455782313, + "grad_norm": 0.9585094451904297, + "learning_rate": 1.4125000000000002e-07, + "loss": 0.0112, + "step": 3887 + }, + { + "epoch": 4.599822537710736, + "grad_norm": 0.8845616579055786, + "learning_rate": 1.4e-07, + "loss": 0.0133, + "step": 3888 + }, + { + "epoch": 4.60100561963916, + "grad_norm": 1.1398485898971558, + "learning_rate": 1.3875000000000002e-07, + "loss": 0.0136, + "step": 3889 + }, + { + "epoch": 4.6021887015675835, + "grad_norm": 1.2279515266418457, + "learning_rate": 1.375e-07, + "loss": 0.0132, + "step": 3890 + }, + { + "epoch": 4.603371783496007, + "grad_norm": 0.8414804935455322, + "learning_rate": 1.3625000000000002e-07, + "loss": 0.0135, + "step": 3891 + }, + { + "epoch": 4.604554865424431, + "grad_norm": 1.300824761390686, + "learning_rate": 1.35e-07, + "loss": 0.012, + "step": 3892 + }, + { + "epoch": 4.605737947352854, + "grad_norm": 0.84298175573349, + "learning_rate": 1.3375000000000002e-07, + "loss": 0.0138, + "step": 3893 + }, + { + "epoch": 4.606921029281278, + "grad_norm": 1.4615131616592407, + "learning_rate": 1.325e-07, + "loss": 0.0135, + "step": 3894 + }, + { + "epoch": 4.608104111209701, + "grad_norm": 1.3707168102264404, + "learning_rate": 1.3125000000000001e-07, + "loss": 0.0155, + "step": 3895 + }, + { + "epoch": 4.609287193138125, + "grad_norm": 0.9600623250007629, + "learning_rate": 1.3e-07, + "loss": 0.0103, + "step": 3896 + }, + { + "epoch": 4.6104702750665485, + "grad_norm": 1.04634428024292, + "learning_rate": 1.2875e-07, + "loss": 0.0108, + "step": 3897 + }, + { + "epoch": 4.611653356994972, + "grad_norm": 1.078597903251648, + "learning_rate": 1.275e-07, + "loss": 0.0121, + "step": 3898 + }, + { + "epoch": 4.612836438923395, + "grad_norm": 1.5998164415359497, + "learning_rate": 1.2625e-07, + "loss": 0.014, + "step": 3899 + }, + { + "epoch": 4.614019520851819, + "grad_norm": 1.631312608718872, + "learning_rate": 1.2500000000000002e-07, + "loss": 0.016, + "step": 3900 + }, + { + "epoch": 4.615202602780243, + "grad_norm": 1.040431022644043, + "learning_rate": 1.2375e-07, + "loss": 0.012, + "step": 3901 + }, + { + "epoch": 4.616385684708666, + "grad_norm": 0.8884280323982239, + "learning_rate": 1.2250000000000002e-07, + "loss": 0.0122, + "step": 3902 + }, + { + "epoch": 4.617568766637089, + "grad_norm": 0.9214581847190857, + "learning_rate": 1.2125e-07, + "loss": 0.0109, + "step": 3903 + }, + { + "epoch": 4.6187518485655135, + "grad_norm": 1.031062364578247, + "learning_rate": 1.2000000000000002e-07, + "loss": 0.0125, + "step": 3904 + }, + { + "epoch": 4.619934930493937, + "grad_norm": 0.8712834119796753, + "learning_rate": 1.1875000000000001e-07, + "loss": 0.0158, + "step": 3905 + }, + { + "epoch": 4.62111801242236, + "grad_norm": 0.994232714176178, + "learning_rate": 1.1750000000000001e-07, + "loss": 0.0114, + "step": 3906 + }, + { + "epoch": 4.622301094350783, + "grad_norm": 1.2968759536743164, + "learning_rate": 1.1625000000000001e-07, + "loss": 0.0125, + "step": 3907 + }, + { + "epoch": 4.623484176279208, + "grad_norm": 1.2021393775939941, + "learning_rate": 1.1500000000000001e-07, + "loss": 0.0133, + "step": 3908 + }, + { + "epoch": 4.624667258207631, + "grad_norm": 1.0258102416992188, + "learning_rate": 1.1375000000000001e-07, + "loss": 0.0129, + "step": 3909 + }, + { + "epoch": 4.625850340136054, + "grad_norm": 1.1024338006973267, + "learning_rate": 1.1250000000000001e-07, + "loss": 0.0114, + "step": 3910 + }, + { + "epoch": 4.627033422064478, + "grad_norm": 1.1806421279907227, + "learning_rate": 1.1125000000000001e-07, + "loss": 0.0124, + "step": 3911 + }, + { + "epoch": 4.628216503992902, + "grad_norm": 1.1528384685516357, + "learning_rate": 1.1e-07, + "loss": 0.0124, + "step": 3912 + }, + { + "epoch": 4.629399585921325, + "grad_norm": 1.1573352813720703, + "learning_rate": 1.0875e-07, + "loss": 0.0139, + "step": 3913 + }, + { + "epoch": 4.630582667849748, + "grad_norm": 0.9932960271835327, + "learning_rate": 1.075e-07, + "loss": 0.0121, + "step": 3914 + }, + { + "epoch": 4.631765749778172, + "grad_norm": 1.2330352067947388, + "learning_rate": 1.0625000000000002e-07, + "loss": 0.016, + "step": 3915 + }, + { + "epoch": 4.632948831706596, + "grad_norm": 1.4231349229812622, + "learning_rate": 1.0500000000000001e-07, + "loss": 0.0118, + "step": 3916 + }, + { + "epoch": 4.634131913635019, + "grad_norm": 0.8875893354415894, + "learning_rate": 1.0375000000000001e-07, + "loss": 0.0142, + "step": 3917 + }, + { + "epoch": 4.6353149955634425, + "grad_norm": 0.9697222113609314, + "learning_rate": 1.0250000000000001e-07, + "loss": 0.0137, + "step": 3918 + }, + { + "epoch": 4.636498077491867, + "grad_norm": 1.2785030603408813, + "learning_rate": 1.0125000000000001e-07, + "loss": 0.0163, + "step": 3919 + }, + { + "epoch": 4.63768115942029, + "grad_norm": 1.0717703104019165, + "learning_rate": 1.0000000000000001e-07, + "loss": 0.012, + "step": 3920 + }, + { + "epoch": 4.638864241348713, + "grad_norm": 1.39628267288208, + "learning_rate": 9.875000000000001e-08, + "loss": 0.0136, + "step": 3921 + }, + { + "epoch": 4.640047323277137, + "grad_norm": 1.1571112871170044, + "learning_rate": 9.75e-08, + "loss": 0.0141, + "step": 3922 + }, + { + "epoch": 4.641230405205561, + "grad_norm": 1.2966231107711792, + "learning_rate": 9.625e-08, + "loss": 0.0135, + "step": 3923 + }, + { + "epoch": 4.642413487133984, + "grad_norm": 1.1383695602416992, + "learning_rate": 9.5e-08, + "loss": 0.0132, + "step": 3924 + }, + { + "epoch": 4.6435965690624075, + "grad_norm": 1.4969112873077393, + "learning_rate": 9.375e-08, + "loss": 0.0162, + "step": 3925 + }, + { + "epoch": 4.644779650990831, + "grad_norm": 1.267492413520813, + "learning_rate": 9.25e-08, + "loss": 0.0153, + "step": 3926 + }, + { + "epoch": 4.645962732919255, + "grad_norm": 1.0656739473342896, + "learning_rate": 9.125e-08, + "loss": 0.0114, + "step": 3927 + }, + { + "epoch": 4.647145814847678, + "grad_norm": 0.9376157522201538, + "learning_rate": 9e-08, + "loss": 0.0113, + "step": 3928 + }, + { + "epoch": 4.648328896776102, + "grad_norm": 1.255050778388977, + "learning_rate": 8.875e-08, + "loss": 0.0134, + "step": 3929 + }, + { + "epoch": 4.649511978704525, + "grad_norm": 1.3018001317977905, + "learning_rate": 8.750000000000001e-08, + "loss": 0.0131, + "step": 3930 + }, + { + "epoch": 4.650695060632949, + "grad_norm": 1.1415715217590332, + "learning_rate": 8.625000000000001e-08, + "loss": 0.0125, + "step": 3931 + }, + { + "epoch": 4.651878142561372, + "grad_norm": 1.0146411657333374, + "learning_rate": 8.500000000000001e-08, + "loss": 0.0133, + "step": 3932 + }, + { + "epoch": 4.653061224489796, + "grad_norm": 0.9796539545059204, + "learning_rate": 8.375e-08, + "loss": 0.0112, + "step": 3933 + }, + { + "epoch": 4.654244306418219, + "grad_norm": 1.529662847518921, + "learning_rate": 8.25e-08, + "loss": 0.0136, + "step": 3934 + }, + { + "epoch": 4.655427388346643, + "grad_norm": 1.138345718383789, + "learning_rate": 8.125e-08, + "loss": 0.0113, + "step": 3935 + }, + { + "epoch": 4.656610470275067, + "grad_norm": 1.33473801612854, + "learning_rate": 8e-08, + "loss": 0.0128, + "step": 3936 + }, + { + "epoch": 4.65779355220349, + "grad_norm": 1.0961664915084839, + "learning_rate": 7.875e-08, + "loss": 0.0142, + "step": 3937 + }, + { + "epoch": 4.658976634131914, + "grad_norm": 0.9592301845550537, + "learning_rate": 7.750000000000001e-08, + "loss": 0.012, + "step": 3938 + }, + { + "epoch": 4.660159716060337, + "grad_norm": 0.990770161151886, + "learning_rate": 7.625000000000001e-08, + "loss": 0.0124, + "step": 3939 + }, + { + "epoch": 4.661342797988761, + "grad_norm": 0.9722172021865845, + "learning_rate": 7.500000000000001e-08, + "loss": 0.0112, + "step": 3940 + }, + { + "epoch": 4.662525879917184, + "grad_norm": 1.5637296438217163, + "learning_rate": 7.375000000000001e-08, + "loss": 0.0156, + "step": 3941 + }, + { + "epoch": 4.663708961845607, + "grad_norm": 0.8520132303237915, + "learning_rate": 7.250000000000001e-08, + "loss": 0.0141, + "step": 3942 + }, + { + "epoch": 4.6648920437740315, + "grad_norm": 1.2372357845306396, + "learning_rate": 7.125e-08, + "loss": 0.0138, + "step": 3943 + }, + { + "epoch": 4.666075125702455, + "grad_norm": 1.3529807329177856, + "learning_rate": 7e-08, + "loss": 0.0134, + "step": 3944 + }, + { + "epoch": 4.667258207630878, + "grad_norm": 1.2668840885162354, + "learning_rate": 6.875e-08, + "loss": 0.0116, + "step": 3945 + }, + { + "epoch": 4.668441289559302, + "grad_norm": 0.9014120101928711, + "learning_rate": 6.75e-08, + "loss": 0.0129, + "step": 3946 + }, + { + "epoch": 4.669624371487726, + "grad_norm": 1.0158212184906006, + "learning_rate": 6.625e-08, + "loss": 0.0132, + "step": 3947 + }, + { + "epoch": 4.670807453416149, + "grad_norm": 1.000591516494751, + "learning_rate": 6.5e-08, + "loss": 0.0109, + "step": 3948 + }, + { + "epoch": 4.671990535344572, + "grad_norm": 1.1380668878555298, + "learning_rate": 6.375e-08, + "loss": 0.0123, + "step": 3949 + }, + { + "epoch": 4.6731736172729965, + "grad_norm": 1.1437523365020752, + "learning_rate": 6.250000000000001e-08, + "loss": 0.014, + "step": 3950 + }, + { + "epoch": 4.67435669920142, + "grad_norm": 1.213784098625183, + "learning_rate": 6.125000000000001e-08, + "loss": 0.0151, + "step": 3951 + }, + { + "epoch": 4.675539781129843, + "grad_norm": 1.241310954093933, + "learning_rate": 6.000000000000001e-08, + "loss": 0.0135, + "step": 3952 + }, + { + "epoch": 4.6767228630582665, + "grad_norm": 1.4583423137664795, + "learning_rate": 5.8750000000000007e-08, + "loss": 0.0122, + "step": 3953 + }, + { + "epoch": 4.677905944986691, + "grad_norm": 1.4477792978286743, + "learning_rate": 5.7500000000000005e-08, + "loss": 0.0119, + "step": 3954 + }, + { + "epoch": 4.679089026915114, + "grad_norm": 1.123445987701416, + "learning_rate": 5.6250000000000004e-08, + "loss": 0.0136, + "step": 3955 + }, + { + "epoch": 4.680272108843537, + "grad_norm": 1.1043845415115356, + "learning_rate": 5.5e-08, + "loss": 0.011, + "step": 3956 + }, + { + "epoch": 4.681455190771961, + "grad_norm": 0.9672202467918396, + "learning_rate": 5.375e-08, + "loss": 0.0144, + "step": 3957 + }, + { + "epoch": 4.682638272700385, + "grad_norm": 1.1119701862335205, + "learning_rate": 5.250000000000001e-08, + "loss": 0.0126, + "step": 3958 + }, + { + "epoch": 4.683821354628808, + "grad_norm": 0.8602145910263062, + "learning_rate": 5.1250000000000006e-08, + "loss": 0.0101, + "step": 3959 + }, + { + "epoch": 4.685004436557231, + "grad_norm": 0.9955289363861084, + "learning_rate": 5.0000000000000004e-08, + "loss": 0.0102, + "step": 3960 + }, + { + "epoch": 4.686187518485655, + "grad_norm": 1.0529160499572754, + "learning_rate": 4.875e-08, + "loss": 0.0123, + "step": 3961 + }, + { + "epoch": 4.687370600414079, + "grad_norm": 0.9052739143371582, + "learning_rate": 4.75e-08, + "loss": 0.013, + "step": 3962 + }, + { + "epoch": 4.688553682342502, + "grad_norm": 1.3269903659820557, + "learning_rate": 4.625e-08, + "loss": 0.0119, + "step": 3963 + }, + { + "epoch": 4.689736764270926, + "grad_norm": 2.333214044570923, + "learning_rate": 4.5e-08, + "loss": 0.0137, + "step": 3964 + }, + { + "epoch": 4.69091984619935, + "grad_norm": 1.2287163734436035, + "learning_rate": 4.3750000000000005e-08, + "loss": 0.0126, + "step": 3965 + }, + { + "epoch": 4.692102928127773, + "grad_norm": 0.9863678216934204, + "learning_rate": 4.2500000000000003e-08, + "loss": 0.0145, + "step": 3966 + }, + { + "epoch": 4.693286010056196, + "grad_norm": 1.2296087741851807, + "learning_rate": 4.125e-08, + "loss": 0.0123, + "step": 3967 + }, + { + "epoch": 4.69446909198462, + "grad_norm": 0.9937695860862732, + "learning_rate": 4e-08, + "loss": 0.0128, + "step": 3968 + }, + { + "epoch": 4.695652173913043, + "grad_norm": 1.1658015251159668, + "learning_rate": 3.8750000000000006e-08, + "loss": 0.0149, + "step": 3969 + }, + { + "epoch": 4.696835255841467, + "grad_norm": 1.1963155269622803, + "learning_rate": 3.7500000000000005e-08, + "loss": 0.0113, + "step": 3970 + }, + { + "epoch": 4.6980183377698905, + "grad_norm": 1.062717318534851, + "learning_rate": 3.6250000000000004e-08, + "loss": 0.0113, + "step": 3971 + }, + { + "epoch": 4.699201419698314, + "grad_norm": 1.2746045589447021, + "learning_rate": 3.5e-08, + "loss": 0.0141, + "step": 3972 + }, + { + "epoch": 4.700384501626738, + "grad_norm": 1.2542712688446045, + "learning_rate": 3.375e-08, + "loss": 0.0147, + "step": 3973 + }, + { + "epoch": 4.701567583555161, + "grad_norm": 0.9702257513999939, + "learning_rate": 3.25e-08, + "loss": 0.0118, + "step": 3974 + }, + { + "epoch": 4.702750665483585, + "grad_norm": 1.3336702585220337, + "learning_rate": 3.1250000000000005e-08, + "loss": 0.0139, + "step": 3975 + }, + { + "epoch": 4.703933747412008, + "grad_norm": 2.4992969036102295, + "learning_rate": 3.0000000000000004e-08, + "loss": 0.0134, + "step": 3976 + }, + { + "epoch": 4.705116829340432, + "grad_norm": 1.003122091293335, + "learning_rate": 2.8750000000000003e-08, + "loss": 0.0144, + "step": 3977 + }, + { + "epoch": 4.7062999112688555, + "grad_norm": 1.039140224456787, + "learning_rate": 2.75e-08, + "loss": 0.0103, + "step": 3978 + }, + { + "epoch": 4.707482993197279, + "grad_norm": 1.1975301504135132, + "learning_rate": 2.6250000000000003e-08, + "loss": 0.0145, + "step": 3979 + }, + { + "epoch": 4.708666075125702, + "grad_norm": 0.960659921169281, + "learning_rate": 2.5000000000000002e-08, + "loss": 0.0091, + "step": 3980 + }, + { + "epoch": 4.709849157054126, + "grad_norm": 1.0543711185455322, + "learning_rate": 2.375e-08, + "loss": 0.0139, + "step": 3981 + }, + { + "epoch": 4.71103223898255, + "grad_norm": 1.198020577430725, + "learning_rate": 2.25e-08, + "loss": 0.0145, + "step": 3982 + }, + { + "epoch": 4.712215320910973, + "grad_norm": 1.0809433460235596, + "learning_rate": 2.1250000000000002e-08, + "loss": 0.0127, + "step": 3983 + }, + { + "epoch": 4.713398402839396, + "grad_norm": 1.449899673461914, + "learning_rate": 2e-08, + "loss": 0.0141, + "step": 3984 + }, + { + "epoch": 4.7145814847678205, + "grad_norm": 1.1461830139160156, + "learning_rate": 1.8750000000000002e-08, + "loss": 0.0137, + "step": 3985 + }, + { + "epoch": 4.715764566696244, + "grad_norm": 1.7491106986999512, + "learning_rate": 1.75e-08, + "loss": 0.0155, + "step": 3986 + }, + { + "epoch": 4.716947648624667, + "grad_norm": 1.956008791923523, + "learning_rate": 1.625e-08, + "loss": 0.015, + "step": 3987 + }, + { + "epoch": 4.71813073055309, + "grad_norm": 1.0659576654434204, + "learning_rate": 1.5000000000000002e-08, + "loss": 0.0108, + "step": 3988 + }, + { + "epoch": 4.719313812481515, + "grad_norm": 1.6309752464294434, + "learning_rate": 1.375e-08, + "loss": 0.0119, + "step": 3989 + }, + { + "epoch": 4.720496894409938, + "grad_norm": 1.061408281326294, + "learning_rate": 1.2500000000000001e-08, + "loss": 0.0148, + "step": 3990 + }, + { + "epoch": 4.721679976338361, + "grad_norm": 1.3093007802963257, + "learning_rate": 1.125e-08, + "loss": 0.0128, + "step": 3991 + }, + { + "epoch": 4.722863058266785, + "grad_norm": 1.16950523853302, + "learning_rate": 1e-08, + "loss": 0.0144, + "step": 3992 + }, + { + "epoch": 4.724046140195209, + "grad_norm": 1.1569727659225464, + "learning_rate": 8.75e-09, + "loss": 0.0124, + "step": 3993 + }, + { + "epoch": 4.725229222123632, + "grad_norm": 0.8837238550186157, + "learning_rate": 7.500000000000001e-09, + "loss": 0.0109, + "step": 3994 + }, + { + "epoch": 4.726412304052055, + "grad_norm": 1.456518530845642, + "learning_rate": 6.2500000000000005e-09, + "loss": 0.012, + "step": 3995 + }, + { + "epoch": 4.727595385980479, + "grad_norm": 0.9517368674278259, + "learning_rate": 5e-09, + "loss": 0.0132, + "step": 3996 + }, + { + "epoch": 4.728778467908903, + "grad_norm": 1.4451509714126587, + "learning_rate": 3.7500000000000005e-09, + "loss": 0.0127, + "step": 3997 + }, + { + "epoch": 4.729961549837326, + "grad_norm": 1.392154335975647, + "learning_rate": 2.5e-09, + "loss": 0.0162, + "step": 3998 + }, + { + "epoch": 4.7311446317657495, + "grad_norm": 1.3502089977264404, + "learning_rate": 1.25e-09, + "loss": 0.0134, + "step": 3999 + }, + { + "epoch": 4.732327713694174, + "grad_norm": 0.9970996379852295, + "learning_rate": 0.0, + "loss": 0.0096, + "step": 4000 + } + ], + "logging_steps": 1.0, + "max_steps": 4000, + "num_input_tokens_seen": 0, + "num_train_epochs": 5, + "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": 9.44568620988498e+19, + "train_batch_size": 4, + "trial_name": null, + "trial_params": null +}