| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 43, | |
| "global_step": 254, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007874015748031496, | |
| "grad_norm": 118.11203002929688, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 4.6099, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007874015748031496, | |
| "eval_loss": 3.1001100540161133, | |
| "eval_runtime": 5.3966, | |
| "eval_samples_per_second": 30.204, | |
| "eval_steps_per_second": 3.891, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.015748031496062992, | |
| "grad_norm": 118.4310302734375, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 4.5857, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.023622047244094488, | |
| "grad_norm": 103.37439727783203, | |
| "learning_rate": 6e-06, | |
| "loss": 4.3069, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.031496062992125984, | |
| "grad_norm": 75.05075073242188, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 3.8754, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03937007874015748, | |
| "grad_norm": 50.459983825683594, | |
| "learning_rate": 1e-05, | |
| "loss": 3.2841, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.047244094488188976, | |
| "grad_norm": 47.4603385925293, | |
| "learning_rate": 1.2e-05, | |
| "loss": 2.4285, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05511811023622047, | |
| "grad_norm": 32.362667083740234, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.8177, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06299212598425197, | |
| "grad_norm": 22.846933364868164, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.1567, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07086614173228346, | |
| "grad_norm": 17.060213088989258, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.8257, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07874015748031496, | |
| "grad_norm": 14.415579795837402, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4257, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08661417322834646, | |
| "grad_norm": 7.753712177276611, | |
| "learning_rate": 1.999964147509006e-05, | |
| "loss": 0.2976, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.09448818897637795, | |
| "grad_norm": 26.883708953857422, | |
| "learning_rate": 1.9998565926068253e-05, | |
| "loss": 0.3365, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.10236220472440945, | |
| "grad_norm": 10.675631523132324, | |
| "learning_rate": 1.9996773430056806e-05, | |
| "loss": 0.2161, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.11023622047244094, | |
| "grad_norm": 6.670111179351807, | |
| "learning_rate": 1.999426411558661e-05, | |
| "loss": 0.1816, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.11811023622047244, | |
| "grad_norm": 8.878239631652832, | |
| "learning_rate": 1.9991038162588018e-05, | |
| "loss": 0.1567, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.12598425196850394, | |
| "grad_norm": 2.9917383193969727, | |
| "learning_rate": 1.9987095802377933e-05, | |
| "loss": 0.0813, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.13385826771653545, | |
| "grad_norm": 1.0548763275146484, | |
| "learning_rate": 1.9982437317643218e-05, | |
| "loss": 0.0217, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.14173228346456693, | |
| "grad_norm": 2.8778488636016846, | |
| "learning_rate": 1.9977063042420438e-05, | |
| "loss": 0.0618, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.14960629921259844, | |
| "grad_norm": 0.9811734557151794, | |
| "learning_rate": 1.99709733620719e-05, | |
| "loss": 0.0175, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.15748031496062992, | |
| "grad_norm": 0.7218202948570251, | |
| "learning_rate": 1.996416871325803e-05, | |
| "loss": 0.0302, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.16535433070866143, | |
| "grad_norm": 1.2746995687484741, | |
| "learning_rate": 1.995664958390604e-05, | |
| "loss": 0.0453, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1732283464566929, | |
| "grad_norm": 0.9413469433784485, | |
| "learning_rate": 1.9948416513174976e-05, | |
| "loss": 0.0175, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.18110236220472442, | |
| "grad_norm": 1.4161137342453003, | |
| "learning_rate": 1.9939470091417012e-05, | |
| "loss": 0.0277, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1889763779527559, | |
| "grad_norm": 2.2721235752105713, | |
| "learning_rate": 1.992981096013517e-05, | |
| "loss": 0.0589, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1968503937007874, | |
| "grad_norm": 1.143970251083374, | |
| "learning_rate": 1.9919439811937283e-05, | |
| "loss": 0.0182, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2047244094488189, | |
| "grad_norm": 0.8054028749465942, | |
| "learning_rate": 1.9908357390486342e-05, | |
| "loss": 0.0211, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2125984251968504, | |
| "grad_norm": 1.4449081420898438, | |
| "learning_rate": 1.989656449044718e-05, | |
| "loss": 0.0244, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2204724409448819, | |
| "grad_norm": 0.49216631054878235, | |
| "learning_rate": 1.988406195742948e-05, | |
| "loss": 0.005, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2283464566929134, | |
| "grad_norm": 0.9945647716522217, | |
| "learning_rate": 1.987085068792715e-05, | |
| "loss": 0.0373, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.23622047244094488, | |
| "grad_norm": 1.1753748655319214, | |
| "learning_rate": 1.9856931629254032e-05, | |
| "loss": 0.0217, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2440944881889764, | |
| "grad_norm": 0.5960403680801392, | |
| "learning_rate": 1.984230577947597e-05, | |
| "loss": 0.0157, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.25196850393700787, | |
| "grad_norm": 0.3657272160053253, | |
| "learning_rate": 1.9826974187339267e-05, | |
| "loss": 0.0082, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.25984251968503935, | |
| "grad_norm": 1.1290266513824463, | |
| "learning_rate": 1.981093795219546e-05, | |
| "loss": 0.0236, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2677165354330709, | |
| "grad_norm": 1.673962116241455, | |
| "learning_rate": 1.9794198223922496e-05, | |
| "loss": 0.0182, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2755905511811024, | |
| "grad_norm": 0.540355384349823, | |
| "learning_rate": 1.9776756202842297e-05, | |
| "loss": 0.011, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.28346456692913385, | |
| "grad_norm": 0.3380790054798126, | |
| "learning_rate": 1.9758613139634662e-05, | |
| "loss": 0.0048, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.29133858267716534, | |
| "grad_norm": 1.886232852935791, | |
| "learning_rate": 1.9739770335247616e-05, | |
| "loss": 0.0157, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2992125984251969, | |
| "grad_norm": 2.140639305114746, | |
| "learning_rate": 1.972022914080411e-05, | |
| "loss": 0.0393, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.30708661417322836, | |
| "grad_norm": 0.35308870673179626, | |
| "learning_rate": 1.9699990957505136e-05, | |
| "loss": 0.0074, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.31496062992125984, | |
| "grad_norm": 0.3918301463127136, | |
| "learning_rate": 1.9679057236529266e-05, | |
| "loss": 0.0083, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3228346456692913, | |
| "grad_norm": 0.4406338632106781, | |
| "learning_rate": 1.965742947892858e-05, | |
| "loss": 0.0152, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.33070866141732286, | |
| "grad_norm": 0.6819682121276855, | |
| "learning_rate": 1.9635109235521057e-05, | |
| "loss": 0.0091, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.33858267716535434, | |
| "grad_norm": 0.6794927716255188, | |
| "learning_rate": 1.961209810677934e-05, | |
| "loss": 0.0071, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.33858267716535434, | |
| "eval_loss": 0.3895845115184784, | |
| "eval_runtime": 6.5602, | |
| "eval_samples_per_second": 24.847, | |
| "eval_steps_per_second": 3.201, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3464566929133858, | |
| "grad_norm": 0.3874967694282532, | |
| "learning_rate": 1.9588397742716004e-05, | |
| "loss": 0.0089, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3543307086614173, | |
| "grad_norm": 0.5577577352523804, | |
| "learning_rate": 1.9564009842765225e-05, | |
| "loss": 0.0098, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.36220472440944884, | |
| "grad_norm": 0.8152347207069397, | |
| "learning_rate": 1.9538936155660934e-05, | |
| "loss": 0.0118, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3700787401574803, | |
| "grad_norm": 0.2971118688583374, | |
| "learning_rate": 1.951317847931141e-05, | |
| "loss": 0.0084, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3779527559055118, | |
| "grad_norm": 1.0286651849746704, | |
| "learning_rate": 1.9486738660670373e-05, | |
| "loss": 0.0123, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.3858267716535433, | |
| "grad_norm": 0.5227222442626953, | |
| "learning_rate": 1.945961859560454e-05, | |
| "loss": 0.0144, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3937007874015748, | |
| "grad_norm": 0.461935818195343, | |
| "learning_rate": 1.943182022875769e-05, | |
| "loss": 0.0119, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4015748031496063, | |
| "grad_norm": 1.2550626993179321, | |
| "learning_rate": 1.940334555341122e-05, | |
| "loss": 0.013, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4094488188976378, | |
| "grad_norm": 0.37549659609794617, | |
| "learning_rate": 1.9374196611341212e-05, | |
| "loss": 0.0181, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.41732283464566927, | |
| "grad_norm": 0.3444191515445709, | |
| "learning_rate": 1.9344375492672024e-05, | |
| "loss": 0.0111, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.4251968503937008, | |
| "grad_norm": 0.3489387333393097, | |
| "learning_rate": 1.9313884335726443e-05, | |
| "loss": 0.0111, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4330708661417323, | |
| "grad_norm": 0.26080814003944397, | |
| "learning_rate": 1.9282725326872324e-05, | |
| "loss": 0.0091, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.4409448818897638, | |
| "grad_norm": 0.1390451341867447, | |
| "learning_rate": 1.9250900700365837e-05, | |
| "loss": 0.0033, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.44881889763779526, | |
| "grad_norm": 0.20499111711978912, | |
| "learning_rate": 1.921841273819125e-05, | |
| "loss": 0.0066, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.4566929133858268, | |
| "grad_norm": 2.185487747192383, | |
| "learning_rate": 1.918526376989731e-05, | |
| "loss": 0.0095, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4645669291338583, | |
| "grad_norm": 0.23939816653728485, | |
| "learning_rate": 1.9151456172430186e-05, | |
| "loss": 0.0048, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.47244094488188976, | |
| "grad_norm": 0.41510018706321716, | |
| "learning_rate": 1.911699236996305e-05, | |
| "loss": 0.0077, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.48031496062992124, | |
| "grad_norm": 0.264318585395813, | |
| "learning_rate": 1.9081874833722234e-05, | |
| "loss": 0.0129, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4881889763779528, | |
| "grad_norm": 1.0443968772888184, | |
| "learning_rate": 1.9046106081810047e-05, | |
| "loss": 0.0035, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.49606299212598426, | |
| "grad_norm": 0.2800132632255554, | |
| "learning_rate": 1.900968867902419e-05, | |
| "loss": 0.0057, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5039370078740157, | |
| "grad_norm": 1.114960789680481, | |
| "learning_rate": 1.8972625236673887e-05, | |
| "loss": 0.0123, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5118110236220472, | |
| "grad_norm": 0.5027065873146057, | |
| "learning_rate": 1.8934918412392596e-05, | |
| "loss": 0.0052, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5196850393700787, | |
| "grad_norm": 0.5564169883728027, | |
| "learning_rate": 1.8896570909947477e-05, | |
| "loss": 0.0085, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5275590551181102, | |
| "grad_norm": 0.7567198872566223, | |
| "learning_rate": 1.8857585479045493e-05, | |
| "loss": 0.0054, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5354330708661418, | |
| "grad_norm": 0.13573969900608063, | |
| "learning_rate": 1.8817964915136277e-05, | |
| "loss": 0.0008, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5433070866141733, | |
| "grad_norm": 0.2704390287399292, | |
| "learning_rate": 1.8777712059211643e-05, | |
| "loss": 0.0078, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5511811023622047, | |
| "grad_norm": 0.6014392971992493, | |
| "learning_rate": 1.8736829797601903e-05, | |
| "loss": 0.0059, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5590551181102362, | |
| "grad_norm": 0.5487034916877747, | |
| "learning_rate": 1.8695321061768886e-05, | |
| "loss": 0.0097, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5669291338582677, | |
| "grad_norm": 0.6670834422111511, | |
| "learning_rate": 1.8653188828095754e-05, | |
| "loss": 0.011, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5748031496062992, | |
| "grad_norm": 0.1795203685760498, | |
| "learning_rate": 1.8610436117673557e-05, | |
| "loss": 0.0067, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5826771653543307, | |
| "grad_norm": 1.768436074256897, | |
| "learning_rate": 1.8567065996084628e-05, | |
| "loss": 0.0096, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5905511811023622, | |
| "grad_norm": 0.26233312487602234, | |
| "learning_rate": 1.8523081573182754e-05, | |
| "loss": 0.0124, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5984251968503937, | |
| "grad_norm": 0.3775719404220581, | |
| "learning_rate": 1.847848600287019e-05, | |
| "loss": 0.0052, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6062992125984252, | |
| "grad_norm": 1.0016565322875977, | |
| "learning_rate": 1.8433282482871497e-05, | |
| "loss": 0.0058, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6141732283464567, | |
| "grad_norm": 0.20153792202472687, | |
| "learning_rate": 1.8387474254504265e-05, | |
| "loss": 0.0056, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6220472440944882, | |
| "grad_norm": 0.5119822025299072, | |
| "learning_rate": 1.8341064602446686e-05, | |
| "loss": 0.0079, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6299212598425197, | |
| "grad_norm": 1.5781004428863525, | |
| "learning_rate": 1.829405685450202e-05, | |
| "loss": 0.008, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6377952755905512, | |
| "grad_norm": 0.23826757073402405, | |
| "learning_rate": 1.824645438135999e-05, | |
| "loss": 0.0041, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.6456692913385826, | |
| "grad_norm": 0.6386727690696716, | |
| "learning_rate": 1.8198260596355077e-05, | |
| "loss": 0.0188, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6535433070866141, | |
| "grad_norm": 0.9503199458122253, | |
| "learning_rate": 1.814947895522176e-05, | |
| "loss": 0.008, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6614173228346457, | |
| "grad_norm": 0.2040701061487198, | |
| "learning_rate": 1.8100112955846746e-05, | |
| "loss": 0.0038, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6692913385826772, | |
| "grad_norm": 0.3660199046134949, | |
| "learning_rate": 1.805016613801813e-05, | |
| "loss": 0.0148, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6771653543307087, | |
| "grad_norm": 1.0502821207046509, | |
| "learning_rate": 1.7999642083171576e-05, | |
| "loss": 0.0098, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6771653543307087, | |
| "eval_loss": 0.3526817262172699, | |
| "eval_runtime": 6.6167, | |
| "eval_samples_per_second": 24.635, | |
| "eval_steps_per_second": 3.174, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6850393700787402, | |
| "grad_norm": 0.13735969364643097, | |
| "learning_rate": 1.7948544414133534e-05, | |
| "loss": 0.0022, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6929133858267716, | |
| "grad_norm": 0.6425012946128845, | |
| "learning_rate": 1.7896876794861443e-05, | |
| "loss": 0.0086, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7007874015748031, | |
| "grad_norm": 0.7540380954742432, | |
| "learning_rate": 1.7844642930181008e-05, | |
| "loss": 0.0062, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7086614173228346, | |
| "grad_norm": 0.6727365255355835, | |
| "learning_rate": 1.779184656552056e-05, | |
| "loss": 0.0027, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7165354330708661, | |
| "grad_norm": 0.14059337973594666, | |
| "learning_rate": 1.773849148664247e-05, | |
| "loss": 0.0056, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.7244094488188977, | |
| "grad_norm": 0.33292093873023987, | |
| "learning_rate": 1.7684581519371714e-05, | |
| "loss": 0.0047, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7322834645669292, | |
| "grad_norm": 0.3809877932071686, | |
| "learning_rate": 1.7630120529321518e-05, | |
| "loss": 0.0139, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.7401574803149606, | |
| "grad_norm": 1.729589819908142, | |
| "learning_rate": 1.7575112421616203e-05, | |
| "loss": 0.0128, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.7480314960629921, | |
| "grad_norm": 0.18192608654499054, | |
| "learning_rate": 1.751956114061113e-05, | |
| "loss": 0.0025, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.7559055118110236, | |
| "grad_norm": 1.0333118438720703, | |
| "learning_rate": 1.7463470669609907e-05, | |
| "loss": 0.006, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.7637795275590551, | |
| "grad_norm": 0.7247685194015503, | |
| "learning_rate": 1.7406845030578747e-05, | |
| "loss": 0.0073, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.7716535433070866, | |
| "grad_norm": 0.06979379802942276, | |
| "learning_rate": 1.734968828385808e-05, | |
| "loss": 0.0005, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.7795275590551181, | |
| "grad_norm": 0.5137119293212891, | |
| "learning_rate": 1.729200452787139e-05, | |
| "loss": 0.0082, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.7874015748031497, | |
| "grad_norm": 0.4704137146472931, | |
| "learning_rate": 1.7233797898831376e-05, | |
| "loss": 0.005, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7952755905511811, | |
| "grad_norm": 0.28564465045928955, | |
| "learning_rate": 1.717507257044331e-05, | |
| "loss": 0.0052, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8031496062992126, | |
| "grad_norm": 0.17685537040233612, | |
| "learning_rate": 1.711583275360582e-05, | |
| "loss": 0.0024, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.8110236220472441, | |
| "grad_norm": 0.45714935660362244, | |
| "learning_rate": 1.7056082696108896e-05, | |
| "loss": 0.0072, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8188976377952756, | |
| "grad_norm": 0.4373086988925934, | |
| "learning_rate": 1.699582668232934e-05, | |
| "loss": 0.0051, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8267716535433071, | |
| "grad_norm": 0.8478983640670776, | |
| "learning_rate": 1.6935069032923525e-05, | |
| "loss": 0.022, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.8346456692913385, | |
| "grad_norm": 0.16181086003780365, | |
| "learning_rate": 1.6873814104517617e-05, | |
| "loss": 0.0058, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.84251968503937, | |
| "grad_norm": 0.09503592550754547, | |
| "learning_rate": 1.6812066289395157e-05, | |
| "loss": 0.0009, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.8503937007874016, | |
| "grad_norm": 0.7462632060050964, | |
| "learning_rate": 1.6749830015182106e-05, | |
| "loss": 0.0044, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.8582677165354331, | |
| "grad_norm": 0.07221701741218567, | |
| "learning_rate": 1.6687109744529394e-05, | |
| "loss": 0.0015, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.8661417322834646, | |
| "grad_norm": 0.08999036252498627, | |
| "learning_rate": 1.6623909974792888e-05, | |
| "loss": 0.0023, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8740157480314961, | |
| "grad_norm": 0.42536938190460205, | |
| "learning_rate": 1.656023523771095e-05, | |
| "loss": 0.005, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.8818897637795275, | |
| "grad_norm": 0.7885191440582275, | |
| "learning_rate": 1.6496090099079452e-05, | |
| "loss": 0.0103, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.889763779527559, | |
| "grad_norm": 0.16610018908977509, | |
| "learning_rate": 1.64314791584244e-05, | |
| "loss": 0.006, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.8976377952755905, | |
| "grad_norm": 0.32151034474372864, | |
| "learning_rate": 1.6366407048672135e-05, | |
| "loss": 0.0086, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.905511811023622, | |
| "grad_norm": 0.557732343673706, | |
| "learning_rate": 1.6300878435817115e-05, | |
| "loss": 0.0064, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9133858267716536, | |
| "grad_norm": 0.2238176167011261, | |
| "learning_rate": 1.6234898018587336e-05, | |
| "loss": 0.0065, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.9212598425196851, | |
| "grad_norm": 0.2980042099952698, | |
| "learning_rate": 1.616847052810744e-05, | |
| "loss": 0.0095, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.9291338582677166, | |
| "grad_norm": 0.1529705822467804, | |
| "learning_rate": 1.6101600727559423e-05, | |
| "loss": 0.0062, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.937007874015748, | |
| "grad_norm": 0.017149658873677254, | |
| "learning_rate": 1.603429341184114e-05, | |
| "loss": 0.0002, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.9448818897637795, | |
| "grad_norm": 0.4514746367931366, | |
| "learning_rate": 1.596655340722244e-05, | |
| "loss": 0.0067, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.952755905511811, | |
| "grad_norm": 0.11766134947538376, | |
| "learning_rate": 1.5898385570999146e-05, | |
| "loss": 0.0053, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.9606299212598425, | |
| "grad_norm": 0.4089784026145935, | |
| "learning_rate": 1.5829794791144723e-05, | |
| "loss": 0.0085, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.968503937007874, | |
| "grad_norm": 0.1353057473897934, | |
| "learning_rate": 1.57607859859598e-05, | |
| "loss": 0.0013, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.9763779527559056, | |
| "grad_norm": 0.6548481583595276, | |
| "learning_rate": 1.5691364103719515e-05, | |
| "loss": 0.0117, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.984251968503937, | |
| "grad_norm": 0.1571267992258072, | |
| "learning_rate": 1.5621534122318682e-05, | |
| "loss": 0.0049, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9921259842519685, | |
| "grad_norm": 1.2177189588546753, | |
| "learning_rate": 1.5551301048914863e-05, | |
| "loss": 0.0161, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.414489209651947, | |
| "learning_rate": 1.5480669919569313e-05, | |
| "loss": 0.0181, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.0078740157480315, | |
| "grad_norm": 0.10985995829105377, | |
| "learning_rate": 1.54096457988859e-05, | |
| "loss": 0.0049, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.015748031496063, | |
| "grad_norm": 0.12780147790908813, | |
| "learning_rate": 1.533823377964791e-05, | |
| "loss": 0.0026, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.015748031496063, | |
| "eval_loss": 0.33064374327659607, | |
| "eval_runtime": 6.9286, | |
| "eval_samples_per_second": 23.526, | |
| "eval_steps_per_second": 3.031, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.0236220472440944, | |
| "grad_norm": 0.5142458081245422, | |
| "learning_rate": 1.52664389824529e-05, | |
| "loss": 0.0082, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.031496062992126, | |
| "grad_norm": 0.15617145597934723, | |
| "learning_rate": 1.5194266555345505e-05, | |
| "loss": 0.0016, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.0393700787401574, | |
| "grad_norm": 0.5782387852668762, | |
| "learning_rate": 1.5121721673448319e-05, | |
| "loss": 0.0117, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.047244094488189, | |
| "grad_norm": 0.08414836972951889, | |
| "learning_rate": 1.5048809538590789e-05, | |
| "loss": 0.0021, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.0551181102362204, | |
| "grad_norm": 0.28253939747810364, | |
| "learning_rate": 1.4975535378936228e-05, | |
| "loss": 0.0055, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.0629921259842519, | |
| "grad_norm": 0.47917842864990234, | |
| "learning_rate": 1.490190444860694e-05, | |
| "loss": 0.0046, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.0708661417322836, | |
| "grad_norm": 0.1895662248134613, | |
| "learning_rate": 1.482792202730745e-05, | |
| "loss": 0.006, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.078740157480315, | |
| "grad_norm": 0.13722768425941467, | |
| "learning_rate": 1.475359341994595e-05, | |
| "loss": 0.0031, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.0866141732283465, | |
| "grad_norm": 0.10731153190135956, | |
| "learning_rate": 1.4678923956253894e-05, | |
| "loss": 0.0005, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.094488188976378, | |
| "grad_norm": 0.12261265516281128, | |
| "learning_rate": 1.460391899040383e-05, | |
| "loss": 0.0031, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.1023622047244095, | |
| "grad_norm": 0.0038245893083512783, | |
| "learning_rate": 1.4528583900625481e-05, | |
| "loss": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.110236220472441, | |
| "grad_norm": 0.28762558102607727, | |
| "learning_rate": 1.4452924088820101e-05, | |
| "loss": 0.004, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.1181102362204725, | |
| "grad_norm": 0.17267552018165588, | |
| "learning_rate": 1.4376944980173138e-05, | |
| "loss": 0.0002, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.125984251968504, | |
| "grad_norm": 0.12727122008800507, | |
| "learning_rate": 1.4300652022765207e-05, | |
| "loss": 0.0029, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.1338582677165354, | |
| "grad_norm": 0.25049135088920593, | |
| "learning_rate": 1.4224050687181442e-05, | |
| "loss": 0.0108, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.141732283464567, | |
| "grad_norm": 0.16092728078365326, | |
| "learning_rate": 1.4147146466119235e-05, | |
| "loss": 0.0024, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.1496062992125984, | |
| "grad_norm": 0.13642658293247223, | |
| "learning_rate": 1.406994487399437e-05, | |
| "loss": 0.0037, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.1574803149606299, | |
| "grad_norm": 0.9029403328895569, | |
| "learning_rate": 1.3992451446545624e-05, | |
| "loss": 0.0034, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.1653543307086613, | |
| "grad_norm": 0.19518424570560455, | |
| "learning_rate": 1.3914671740437811e-05, | |
| "loss": 0.0057, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.1732283464566928, | |
| "grad_norm": 0.12140502035617828, | |
| "learning_rate": 1.3836611332863356e-05, | |
| "loss": 0.0041, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.1811023622047245, | |
| "grad_norm": 0.5148038864135742, | |
| "learning_rate": 1.3758275821142382e-05, | |
| "loss": 0.0026, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.188976377952756, | |
| "grad_norm": 1.828904390335083, | |
| "learning_rate": 1.3679670822321347e-05, | |
| "loss": 0.0024, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.1968503937007875, | |
| "grad_norm": 0.3571717143058777, | |
| "learning_rate": 1.3600801972770272e-05, | |
| "loss": 0.0106, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.204724409448819, | |
| "grad_norm": 0.051027003675699234, | |
| "learning_rate": 1.3521674927778594e-05, | |
| "loss": 0.0003, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.2125984251968505, | |
| "grad_norm": 0.6490982174873352, | |
| "learning_rate": 1.3442295361149651e-05, | |
| "loss": 0.0035, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.220472440944882, | |
| "grad_norm": 0.08408445864915848, | |
| "learning_rate": 1.336266896479384e-05, | |
| "loss": 0.0027, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.2283464566929134, | |
| "grad_norm": 0.09666562080383301, | |
| "learning_rate": 1.328280144832047e-05, | |
| "loss": 0.0019, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.236220472440945, | |
| "grad_norm": 0.03880690038204193, | |
| "learning_rate": 1.3202698538628376e-05, | |
| "loss": 0.0003, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.2440944881889764, | |
| "grad_norm": 0.11940775066614151, | |
| "learning_rate": 1.3122365979495259e-05, | |
| "loss": 0.0024, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.2519685039370079, | |
| "grad_norm": 0.1442880481481552, | |
| "learning_rate": 1.3041809531165819e-05, | |
| "loss": 0.0015, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.2598425196850394, | |
| "grad_norm": 0.1961939036846161, | |
| "learning_rate": 1.2961034969938732e-05, | |
| "loss": 0.0056, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.2677165354330708, | |
| "grad_norm": 0.26947638392448425, | |
| "learning_rate": 1.288004808775246e-05, | |
| "loss": 0.0028, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.2755905511811023, | |
| "grad_norm": 0.5154056549072266, | |
| "learning_rate": 1.2798854691769927e-05, | |
| "loss": 0.0037, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.2834645669291338, | |
| "grad_norm": 0.4292369782924652, | |
| "learning_rate": 1.2717460603962132e-05, | |
| "loss": 0.0029, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.2913385826771653, | |
| "grad_norm": 0.19139212369918823, | |
| "learning_rate": 1.2635871660690677e-05, | |
| "loss": 0.0061, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.2992125984251968, | |
| "grad_norm": 0.19960306584835052, | |
| "learning_rate": 1.2554093712289267e-05, | |
| "loss": 0.005, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.3070866141732282, | |
| "grad_norm": 0.4523830711841583, | |
| "learning_rate": 1.2472132622644222e-05, | |
| "loss": 0.0065, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.3149606299212597, | |
| "grad_norm": 0.49343299865722656, | |
| "learning_rate": 1.2389994268773995e-05, | |
| "loss": 0.0061, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.3228346456692912, | |
| "grad_norm": 0.01938088797032833, | |
| "learning_rate": 1.2307684540407775e-05, | |
| "loss": 0.0001, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.330708661417323, | |
| "grad_norm": 0.3082112669944763, | |
| "learning_rate": 1.2225209339563144e-05, | |
| "loss": 0.0053, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.3385826771653544, | |
| "grad_norm": 0.01982509344816208, | |
| "learning_rate": 1.2142574580122903e-05, | |
| "loss": 0.0001, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3464566929133859, | |
| "grad_norm": 0.12388588488101959, | |
| "learning_rate": 1.2059786187410984e-05, | |
| "loss": 0.0049, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.3543307086614174, | |
| "grad_norm": 0.43759095668792725, | |
| "learning_rate": 1.1976850097767598e-05, | |
| "loss": 0.0128, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.3543307086614174, | |
| "eval_loss": 0.3166251480579376, | |
| "eval_runtime": 6.9515, | |
| "eval_samples_per_second": 23.448, | |
| "eval_steps_per_second": 3.021, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.3622047244094488, | |
| "grad_norm": 0.46561670303344727, | |
| "learning_rate": 1.1893772258123554e-05, | |
| "loss": 0.008, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.3700787401574803, | |
| "grad_norm": 0.16612188518047333, | |
| "learning_rate": 1.1810558625573856e-05, | |
| "loss": 0.0024, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.3779527559055118, | |
| "grad_norm": 0.13628093898296356, | |
| "learning_rate": 1.1727215166950519e-05, | |
| "loss": 0.0045, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.3858267716535433, | |
| "grad_norm": 0.565229058265686, | |
| "learning_rate": 1.1643747858394743e-05, | |
| "loss": 0.0103, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.3937007874015748, | |
| "grad_norm": 0.14550763368606567, | |
| "learning_rate": 1.156016268492839e-05, | |
| "loss": 0.0028, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.4015748031496063, | |
| "grad_norm": 0.12460129708051682, | |
| "learning_rate": 1.1476465640024814e-05, | |
| "loss": 0.0031, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.4094488188976377, | |
| "grad_norm": 0.19089221954345703, | |
| "learning_rate": 1.1392662725179114e-05, | |
| "loss": 0.0035, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.4173228346456692, | |
| "grad_norm": 0.6106573343276978, | |
| "learning_rate": 1.1308759949477786e-05, | |
| "loss": 0.0088, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.425196850393701, | |
| "grad_norm": 0.20053207874298096, | |
| "learning_rate": 1.1224763329167859e-05, | |
| "loss": 0.0033, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.4330708661417324, | |
| "grad_norm": 0.1984691321849823, | |
| "learning_rate": 1.1140678887225468e-05, | |
| "loss": 0.0051, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.4409448818897639, | |
| "grad_norm": 0.19264858961105347, | |
| "learning_rate": 1.1056512652924014e-05, | |
| "loss": 0.0046, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.4488188976377954, | |
| "grad_norm": 0.10979076474905014, | |
| "learning_rate": 1.0972270661401812e-05, | |
| "loss": 0.0031, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.4566929133858268, | |
| "grad_norm": 0.1744084656238556, | |
| "learning_rate": 1.0887958953229349e-05, | |
| "loss": 0.0024, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.4645669291338583, | |
| "grad_norm": 0.20646224915981293, | |
| "learning_rate": 1.0803583573976137e-05, | |
| "loss": 0.008, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.4724409448818898, | |
| "grad_norm": 0.14391584694385529, | |
| "learning_rate": 1.0719150573777226e-05, | |
| "loss": 0.004, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.4803149606299213, | |
| "grad_norm": 0.36887863278388977, | |
| "learning_rate": 1.0634666006899375e-05, | |
| "loss": 0.0074, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.4881889763779528, | |
| "grad_norm": 0.21352627873420715, | |
| "learning_rate": 1.055013593130693e-05, | |
| "loss": 0.0082, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.4960629921259843, | |
| "grad_norm": 0.22443020343780518, | |
| "learning_rate": 1.046556640822744e-05, | |
| "loss": 0.0087, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.5039370078740157, | |
| "grad_norm": 0.4243764281272888, | |
| "learning_rate": 1.0380963501717034e-05, | |
| "loss": 0.0068, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.5118110236220472, | |
| "grad_norm": 0.17558562755584717, | |
| "learning_rate": 1.0296333278225599e-05, | |
| "loss": 0.0054, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.5196850393700787, | |
| "grad_norm": 0.14842620491981506, | |
| "learning_rate": 1.0211681806161787e-05, | |
| "loss": 0.0031, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.5275590551181102, | |
| "grad_norm": 0.09316081553697586, | |
| "learning_rate": 1.0127015155457875e-05, | |
| "loss": 0.0013, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.5354330708661417, | |
| "grad_norm": 0.19795025885105133, | |
| "learning_rate": 1.0042339397134528e-05, | |
| "loss": 0.0051, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.5433070866141732, | |
| "grad_norm": 0.21606990694999695, | |
| "learning_rate": 9.957660602865477e-06, | |
| "loss": 0.0041, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.5511811023622046, | |
| "grad_norm": 0.18036173284053802, | |
| "learning_rate": 9.872984844542128e-06, | |
| "loss": 0.0037, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.5590551181102361, | |
| "grad_norm": 0.18953870236873627, | |
| "learning_rate": 9.788318193838218e-06, | |
| "loss": 0.0041, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.5669291338582676, | |
| "grad_norm": 0.12346503138542175, | |
| "learning_rate": 9.703666721774403e-06, | |
| "loss": 0.0035, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.574803149606299, | |
| "grad_norm": 0.4576225280761719, | |
| "learning_rate": 9.619036498282968e-06, | |
| "loss": 0.0041, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.5826771653543306, | |
| "grad_norm": 0.10333681106567383, | |
| "learning_rate": 9.534433591772562e-06, | |
| "loss": 0.0011, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.590551181102362, | |
| "grad_norm": 0.19167865812778473, | |
| "learning_rate": 9.449864068693072e-06, | |
| "loss": 0.0062, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.5984251968503937, | |
| "grad_norm": 0.2258184254169464, | |
| "learning_rate": 9.365333993100628e-06, | |
| "loss": 0.003, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.6062992125984252, | |
| "grad_norm": 0.07945302873849869, | |
| "learning_rate": 9.280849426222778e-06, | |
| "loss": 0.0008, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.6141732283464567, | |
| "grad_norm": 0.17767398059368134, | |
| "learning_rate": 9.196416426023868e-06, | |
| "loss": 0.0053, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.6220472440944882, | |
| "grad_norm": 0.12704500555992126, | |
| "learning_rate": 9.112041046770653e-06, | |
| "loss": 0.0023, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.6299212598425197, | |
| "grad_norm": 0.4054742753505707, | |
| "learning_rate": 9.027729338598188e-06, | |
| "loss": 0.0045, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.6377952755905512, | |
| "grad_norm": 0.4463757574558258, | |
| "learning_rate": 8.943487347075988e-06, | |
| "loss": 0.007, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.6456692913385826, | |
| "grad_norm": 0.6517045497894287, | |
| "learning_rate": 8.859321112774535e-06, | |
| "loss": 0.0052, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.6535433070866141, | |
| "grad_norm": 0.1542089730501175, | |
| "learning_rate": 8.775236670832146e-06, | |
| "loss": 0.0047, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.6614173228346458, | |
| "grad_norm": 0.14716440439224243, | |
| "learning_rate": 8.691240050522215e-06, | |
| "loss": 0.0049, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.6692913385826773, | |
| "grad_norm": 0.2997347116470337, | |
| "learning_rate": 8.607337274820888e-06, | |
| "loss": 0.0076, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.6771653543307088, | |
| "grad_norm": 0.22548256814479828, | |
| "learning_rate": 8.52353435997519e-06, | |
| "loss": 0.0063, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.6850393700787403, | |
| "grad_norm": 0.7220733165740967, | |
| "learning_rate": 8.439837315071612e-06, | |
| "loss": 0.0089, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.6929133858267718, | |
| "grad_norm": 0.5101618766784668, | |
| "learning_rate": 8.35625214160526e-06, | |
| "loss": 0.0042, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.6929133858267718, | |
| "eval_loss": 0.3484288156032562, | |
| "eval_runtime": 6.4482, | |
| "eval_samples_per_second": 25.278, | |
| "eval_steps_per_second": 3.257, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.7007874015748032, | |
| "grad_norm": 0.1698393076658249, | |
| "learning_rate": 8.272784833049485e-06, | |
| "loss": 0.0028, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.7086614173228347, | |
| "grad_norm": 0.5772718191146851, | |
| "learning_rate": 8.18944137442615e-06, | |
| "loss": 0.0082, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.7165354330708662, | |
| "grad_norm": 0.09606469422578812, | |
| "learning_rate": 8.106227741876447e-06, | |
| "loss": 0.0011, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.7244094488188977, | |
| "grad_norm": 0.14510361850261688, | |
| "learning_rate": 8.023149902232404e-06, | |
| "loss": 0.0015, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.7322834645669292, | |
| "grad_norm": 0.055804118514060974, | |
| "learning_rate": 7.940213812589018e-06, | |
| "loss": 0.0008, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.7401574803149606, | |
| "grad_norm": 0.13318321108818054, | |
| "learning_rate": 7.857425419877097e-06, | |
| "loss": 0.005, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.7480314960629921, | |
| "grad_norm": 0.23600782454013824, | |
| "learning_rate": 7.774790660436857e-06, | |
| "loss": 0.0063, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.7559055118110236, | |
| "grad_norm": 0.8483791351318359, | |
| "learning_rate": 7.69231545959223e-06, | |
| "loss": 0.0027, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.763779527559055, | |
| "grad_norm": 0.16536197066307068, | |
| "learning_rate": 7.610005731226009e-06, | |
| "loss": 0.0039, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.7716535433070866, | |
| "grad_norm": 0.14446765184402466, | |
| "learning_rate": 7.52786737735578e-06, | |
| "loss": 0.0036, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.779527559055118, | |
| "grad_norm": 0.8880365490913391, | |
| "learning_rate": 7.445906287710733e-06, | |
| "loss": 0.0061, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.7874015748031495, | |
| "grad_norm": 0.151743084192276, | |
| "learning_rate": 7.364128339309326e-06, | |
| "loss": 0.0028, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.795275590551181, | |
| "grad_norm": 0.1224551647901535, | |
| "learning_rate": 7.282539396037868e-06, | |
| "loss": 0.002, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.8031496062992125, | |
| "grad_norm": 0.4868486225605011, | |
| "learning_rate": 7.201145308230075e-06, | |
| "loss": 0.0031, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.811023622047244, | |
| "grad_norm": 0.2875569462776184, | |
| "learning_rate": 7.119951912247545e-06, | |
| "loss": 0.0082, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.8188976377952755, | |
| "grad_norm": 0.43524420261383057, | |
| "learning_rate": 7.038965030061273e-06, | |
| "loss": 0.0075, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.826771653543307, | |
| "grad_norm": 0.39634883403778076, | |
| "learning_rate": 6.9581904688341854e-06, | |
| "loss": 0.0032, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.8346456692913384, | |
| "grad_norm": 0.9809433817863464, | |
| "learning_rate": 6.8776340205047446e-06, | |
| "loss": 0.0085, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.84251968503937, | |
| "grad_norm": 0.20062875747680664, | |
| "learning_rate": 6.797301461371626e-06, | |
| "loss": 0.0043, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.8503937007874016, | |
| "grad_norm": 0.148948073387146, | |
| "learning_rate": 6.7171985516795315e-06, | |
| "loss": 0.0036, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.858267716535433, | |
| "grad_norm": 0.15658679604530334, | |
| "learning_rate": 6.637331035206166e-06, | |
| "loss": 0.0046, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.8661417322834646, | |
| "grad_norm": 0.22365815937519073, | |
| "learning_rate": 6.557704638850352e-06, | |
| "loss": 0.0081, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.874015748031496, | |
| "grad_norm": 0.10596666485071182, | |
| "learning_rate": 6.4783250722214066e-06, | |
| "loss": 0.0032, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.8818897637795275, | |
| "grad_norm": 0.2130754142999649, | |
| "learning_rate": 6.399198027229732e-06, | |
| "loss": 0.0056, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.889763779527559, | |
| "grad_norm": 0.05641167238354683, | |
| "learning_rate": 6.320329177678656e-06, | |
| "loss": 0.0008, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.8976377952755905, | |
| "grad_norm": 0.10349344462156296, | |
| "learning_rate": 6.241724178857621e-06, | |
| "loss": 0.0026, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.905511811023622, | |
| "grad_norm": 0.08451675623655319, | |
| "learning_rate": 6.163388667136646e-06, | |
| "loss": 0.0016, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.9133858267716537, | |
| "grad_norm": 0.13671623170375824, | |
| "learning_rate": 6.085328259562195e-06, | |
| "loss": 0.0034, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.9212598425196852, | |
| "grad_norm": 0.5500523447990417, | |
| "learning_rate": 6.007548553454379e-06, | |
| "loss": 0.0028, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.9291338582677167, | |
| "grad_norm": 0.06702329218387604, | |
| "learning_rate": 5.93005512600563e-06, | |
| "loss": 0.0009, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.9370078740157481, | |
| "grad_norm": 0.15156973898410797, | |
| "learning_rate": 5.852853533880768e-06, | |
| "loss": 0.0064, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.9448818897637796, | |
| "grad_norm": 0.2970314621925354, | |
| "learning_rate": 5.7759493128185584e-06, | |
| "loss": 0.0077, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.952755905511811, | |
| "grad_norm": 0.06406261771917343, | |
| "learning_rate": 5.699347977234799e-06, | |
| "loss": 0.0006, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.9606299212598426, | |
| "grad_norm": 0.2910393178462982, | |
| "learning_rate": 5.623055019826862e-06, | |
| "loss": 0.0036, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.968503937007874, | |
| "grad_norm": 0.6454993486404419, | |
| "learning_rate": 5.547075911179902e-06, | |
| "loss": 0.0084, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.9763779527559056, | |
| "grad_norm": 0.09460143744945526, | |
| "learning_rate": 5.471416099374525e-06, | |
| "loss": 0.0021, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.984251968503937, | |
| "grad_norm": 0.2024363875389099, | |
| "learning_rate": 5.3960810095961705e-06, | |
| "loss": 0.0052, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.9921259842519685, | |
| "grad_norm": 0.09423142671585083, | |
| "learning_rate": 5.321076043746108e-06, | |
| "loss": 0.0018, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.1085880920290947, | |
| "learning_rate": 5.246406580054051e-06, | |
| "loss": 0.0039, | |
| "step": 254 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 381, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 127, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.6174542139778662e+17, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |