| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.987654320987654, | |
| "eval_steps": 500, | |
| "global_step": 225, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02194787379972565, | |
| "grad_norm": 6.221677266588304, | |
| "learning_rate": 3.4782608695652175e-06, | |
| "loss": 0.8718, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0438957475994513, | |
| "grad_norm": 6.253264896473354, | |
| "learning_rate": 6.956521739130435e-06, | |
| "loss": 0.8766, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06584362139917696, | |
| "grad_norm": 5.741674847301876, | |
| "learning_rate": 1.0434782608695653e-05, | |
| "loss": 0.8549, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0877914951989026, | |
| "grad_norm": 2.3877482149850215, | |
| "learning_rate": 1.391304347826087e-05, | |
| "loss": 0.7607, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.10973936899862825, | |
| "grad_norm": 4.563966782136464, | |
| "learning_rate": 1.739130434782609e-05, | |
| "loss": 0.7707, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.13168724279835392, | |
| "grad_norm": 5.736100045190483, | |
| "learning_rate": 2.0869565217391306e-05, | |
| "loss": 0.7229, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.15363511659807957, | |
| "grad_norm": 5.958244309479874, | |
| "learning_rate": 2.4347826086956526e-05, | |
| "loss": 0.7388, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1755829903978052, | |
| "grad_norm": 2.9204724652960246, | |
| "learning_rate": 2.782608695652174e-05, | |
| "loss": 0.6832, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 2.3986017638483466, | |
| "learning_rate": 3.130434782608696e-05, | |
| "loss": 0.649, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2194787379972565, | |
| "grad_norm": 1.6178731632508974, | |
| "learning_rate": 3.478260869565218e-05, | |
| "loss": 0.6185, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.24142661179698216, | |
| "grad_norm": 1.6322225805622788, | |
| "learning_rate": 3.8260869565217395e-05, | |
| "loss": 0.6045, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.26337448559670784, | |
| "grad_norm": 1.538396522627224, | |
| "learning_rate": 4.173913043478261e-05, | |
| "loss": 0.5934, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.28532235939643347, | |
| "grad_norm": 1.765556579505549, | |
| "learning_rate": 4.521739130434783e-05, | |
| "loss": 0.6012, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.30727023319615915, | |
| "grad_norm": 1.9039962720526176, | |
| "learning_rate": 4.869565217391305e-05, | |
| "loss": 0.5794, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.3292181069958848, | |
| "grad_norm": 0.9685232076545514, | |
| "learning_rate": 5.217391304347826e-05, | |
| "loss": 0.5783, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3511659807956104, | |
| "grad_norm": 2.3819790183508056, | |
| "learning_rate": 5.565217391304348e-05, | |
| "loss": 0.586, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3731138545953361, | |
| "grad_norm": 1.4218859481860902, | |
| "learning_rate": 5.91304347826087e-05, | |
| "loss": 0.5584, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 1.705949751999593, | |
| "learning_rate": 6.260869565217392e-05, | |
| "loss": 0.5785, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.4170096021947874, | |
| "grad_norm": 1.4612385852414467, | |
| "learning_rate": 6.608695652173914e-05, | |
| "loss": 0.562, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.438957475994513, | |
| "grad_norm": 0.9429650218903195, | |
| "learning_rate": 6.956521739130436e-05, | |
| "loss": 0.5586, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4609053497942387, | |
| "grad_norm": 1.8504062874353713, | |
| "learning_rate": 7.304347826086957e-05, | |
| "loss": 0.5648, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.4828532235939643, | |
| "grad_norm": 1.0359118447076778, | |
| "learning_rate": 7.652173913043479e-05, | |
| "loss": 0.5448, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.50480109739369, | |
| "grad_norm": 1.6864037499836835, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5592, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.5267489711934157, | |
| "grad_norm": 1.2434910951524751, | |
| "learning_rate": 7.999516253040599e-05, | |
| "loss": 0.5427, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5486968449931413, | |
| "grad_norm": 1.3000636051596768, | |
| "learning_rate": 7.998065129167953e-05, | |
| "loss": 0.5373, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5706447187928669, | |
| "grad_norm": 2.684317103873129, | |
| "learning_rate": 7.995646979370445e-05, | |
| "loss": 0.5399, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 1.4021363857201035, | |
| "learning_rate": 7.992262388534378e-05, | |
| "loss": 0.5434, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.6145404663923183, | |
| "grad_norm": 1.9180218278821586, | |
| "learning_rate": 7.987912175302518e-05, | |
| "loss": 0.5344, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.6364883401920439, | |
| "grad_norm": 1.152289451077819, | |
| "learning_rate": 7.982597391876076e-05, | |
| "loss": 0.531, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6584362139917695, | |
| "grad_norm": 1.8585951370881602, | |
| "learning_rate": 7.976319323760211e-05, | |
| "loss": 0.5416, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6803840877914952, | |
| "grad_norm": 1.3000646333049437, | |
| "learning_rate": 7.969079489453107e-05, | |
| "loss": 0.5263, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.7023319615912208, | |
| "grad_norm": 1.6442425191809935, | |
| "learning_rate": 7.960879640078679e-05, | |
| "loss": 0.5282, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.7242798353909465, | |
| "grad_norm": 1.3282337661300934, | |
| "learning_rate": 7.951721758963028e-05, | |
| "loss": 0.5125, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.7462277091906722, | |
| "grad_norm": 1.1249842886051156, | |
| "learning_rate": 7.941608061154726e-05, | |
| "loss": 0.5066, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7681755829903978, | |
| "grad_norm": 0.9153543962795861, | |
| "learning_rate": 7.930540992889056e-05, | |
| "loss": 0.5041, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.8714948501569295, | |
| "learning_rate": 7.918523230996327e-05, | |
| "loss": 0.5078, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.8120713305898491, | |
| "grad_norm": 1.17614468677297, | |
| "learning_rate": 7.905557682254429e-05, | |
| "loss": 0.5132, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.8340192043895748, | |
| "grad_norm": 0.9070928796685083, | |
| "learning_rate": 7.891647482685751e-05, | |
| "loss": 0.4954, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.8559670781893004, | |
| "grad_norm": 0.6348754426964612, | |
| "learning_rate": 7.876795996798665e-05, | |
| "loss": 0.5002, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.877914951989026, | |
| "grad_norm": 0.7240683284862476, | |
| "learning_rate": 7.861006816773743e-05, | |
| "loss": 0.4974, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8998628257887518, | |
| "grad_norm": 0.6967632259977047, | |
| "learning_rate": 7.844283761594899e-05, | |
| "loss": 0.4916, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.9218106995884774, | |
| "grad_norm": 0.7523038917826653, | |
| "learning_rate": 7.82663087612568e-05, | |
| "loss": 0.4943, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.943758573388203, | |
| "grad_norm": 0.9009479018151323, | |
| "learning_rate": 7.80805243013092e-05, | |
| "loss": 0.4948, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.9657064471879286, | |
| "grad_norm": 1.1416188350869176, | |
| "learning_rate": 7.788552917244002e-05, | |
| "loss": 0.4963, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.7724926247499572, | |
| "learning_rate": 7.768137053879957e-05, | |
| "loss": 0.4827, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.00960219478738, | |
| "grad_norm": 0.7378875298449712, | |
| "learning_rate": 7.7468097780947e-05, | |
| "loss": 0.4808, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.0315500685871055, | |
| "grad_norm": 0.8154669296425998, | |
| "learning_rate": 7.724576248390639e-05, | |
| "loss": 0.4702, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.0534979423868314, | |
| "grad_norm": 0.7479771108510165, | |
| "learning_rate": 7.701441842468968e-05, | |
| "loss": 0.457, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.075445816186557, | |
| "grad_norm": 0.8109214505078363, | |
| "learning_rate": 7.677412155928946e-05, | |
| "loss": 0.4581, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.0973936899862826, | |
| "grad_norm": 0.620878499140312, | |
| "learning_rate": 7.652493000914476e-05, | |
| "loss": 0.4643, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.1193415637860082, | |
| "grad_norm": 0.41252079635226235, | |
| "learning_rate": 7.62669040470829e-05, | |
| "loss": 0.4643, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.1412894375857339, | |
| "grad_norm": 0.512997391865451, | |
| "learning_rate": 7.60001060827412e-05, | |
| "loss": 0.4566, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.1632373113854595, | |
| "grad_norm": 0.5520989749480719, | |
| "learning_rate": 7.572460064747167e-05, | |
| "loss": 0.4553, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.48682230835165313, | |
| "learning_rate": 7.544045437873259e-05, | |
| "loss": 0.4503, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.2071330589849107, | |
| "grad_norm": 0.40890214708757283, | |
| "learning_rate": 7.514773600397076e-05, | |
| "loss": 0.449, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.2290809327846364, | |
| "grad_norm": 0.4496051089290013, | |
| "learning_rate": 7.484651632399802e-05, | |
| "loss": 0.4486, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.2510288065843622, | |
| "grad_norm": 0.3385328491192872, | |
| "learning_rate": 7.453686819586655e-05, | |
| "loss": 0.4473, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.2729766803840878, | |
| "grad_norm": 0.5042801234415435, | |
| "learning_rate": 7.421886651524656e-05, | |
| "loss": 0.4419, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.2949245541838135, | |
| "grad_norm": 0.4013529469572051, | |
| "learning_rate": 7.38925881983111e-05, | |
| "loss": 0.4463, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.316872427983539, | |
| "grad_norm": 0.3526194553768413, | |
| "learning_rate": 7.355811216313206e-05, | |
| "loss": 0.4446, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.3388203017832647, | |
| "grad_norm": 0.3798310275778352, | |
| "learning_rate": 7.321551931059191e-05, | |
| "loss": 0.4548, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.3607681755829903, | |
| "grad_norm": 0.4050313385315871, | |
| "learning_rate": 7.286489250481604e-05, | |
| "loss": 0.4473, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.5307566330866562, | |
| "learning_rate": 7.250631655313001e-05, | |
| "loss": 0.4428, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.4046639231824418, | |
| "grad_norm": 0.5886517904245452, | |
| "learning_rate": 7.213987818554704e-05, | |
| "loss": 0.4439, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.4266117969821672, | |
| "grad_norm": 0.7870712747634117, | |
| "learning_rate": 7.176566603379015e-05, | |
| "loss": 0.4553, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.448559670781893, | |
| "grad_norm": 1.0045619121977905, | |
| "learning_rate": 7.138377060985465e-05, | |
| "loss": 0.4467, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.4705075445816187, | |
| "grad_norm": 1.0300458801690289, | |
| "learning_rate": 7.09942842841156e-05, | |
| "loss": 0.4481, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.4924554183813443, | |
| "grad_norm": 0.6788257591381656, | |
| "learning_rate": 7.059730126298591e-05, | |
| "loss": 0.4446, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.51440329218107, | |
| "grad_norm": 0.5581173116767874, | |
| "learning_rate": 7.019291756613029e-05, | |
| "loss": 0.4397, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.5363511659807956, | |
| "grad_norm": 0.740704532124767, | |
| "learning_rate": 6.978123100324061e-05, | |
| "loss": 0.4402, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.5582990397805214, | |
| "grad_norm": 0.7781231979915648, | |
| "learning_rate": 6.936234115037842e-05, | |
| "loss": 0.4429, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.6101020638409748, | |
| "learning_rate": 6.893634932589e-05, | |
| "loss": 0.4433, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.6021947873799727, | |
| "grad_norm": 0.6338457181714169, | |
| "learning_rate": 6.85033585659003e-05, | |
| "loss": 0.4389, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.624142661179698, | |
| "grad_norm": 0.6492977834572081, | |
| "learning_rate": 6.806347359939107e-05, | |
| "loss": 0.4351, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.646090534979424, | |
| "grad_norm": 0.392861233479167, | |
| "learning_rate": 6.761680082286988e-05, | |
| "loss": 0.4399, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.6680384087791496, | |
| "grad_norm": 0.4720757546111325, | |
| "learning_rate": 6.716344827463545e-05, | |
| "loss": 0.4349, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.6899862825788752, | |
| "grad_norm": 0.518992673759433, | |
| "learning_rate": 6.670352560864615e-05, | |
| "loss": 0.4384, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.7119341563786008, | |
| "grad_norm": 0.3064648926060582, | |
| "learning_rate": 6.62371440679976e-05, | |
| "loss": 0.4298, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.7338820301783264, | |
| "grad_norm": 0.2952380631214821, | |
| "learning_rate": 6.576441645801592e-05, | |
| "loss": 0.4384, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.7558299039780523, | |
| "grad_norm": 0.3619917164563617, | |
| "learning_rate": 6.528545711897307e-05, | |
| "loss": 0.4367, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.3352348275889567, | |
| "learning_rate": 6.480038189843101e-05, | |
| "loss": 0.4411, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.7997256515775035, | |
| "grad_norm": 0.3510930618032657, | |
| "learning_rate": 6.430930812322127e-05, | |
| "loss": 0.4386, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.821673525377229, | |
| "grad_norm": 0.4010609683995807, | |
| "learning_rate": 6.381235457106664e-05, | |
| "loss": 0.4298, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.8436213991769548, | |
| "grad_norm": 0.36451457683299865, | |
| "learning_rate": 6.330964144185204e-05, | |
| "loss": 0.4371, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.8655692729766804, | |
| "grad_norm": 0.27838534336167414, | |
| "learning_rate": 6.280129032855132e-05, | |
| "loss": 0.4305, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.887517146776406, | |
| "grad_norm": 0.2966462979039974, | |
| "learning_rate": 6.228742418781714e-05, | |
| "loss": 0.4388, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.9094650205761317, | |
| "grad_norm": 0.26944878326607974, | |
| "learning_rate": 6.17681673102411e-05, | |
| "loss": 0.4345, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.9314128943758573, | |
| "grad_norm": 0.2712568832140498, | |
| "learning_rate": 6.1243645290291e-05, | |
| "loss": 0.4308, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.9533607681755831, | |
| "grad_norm": 0.2866166967589882, | |
| "learning_rate": 6.0713984995933016e-05, | |
| "loss": 0.4325, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.23491982047748894, | |
| "learning_rate": 6.01793145379456e-05, | |
| "loss": 0.4321, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.9972565157750344, | |
| "grad_norm": 0.205586198763508, | |
| "learning_rate": 5.9639763238932893e-05, | |
| "loss": 0.4345, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.01920438957476, | |
| "grad_norm": 0.27481178105933207, | |
| "learning_rate": 5.909546160204508e-05, | |
| "loss": 0.4013, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.0411522633744856, | |
| "grad_norm": 0.30806418646748673, | |
| "learning_rate": 5.8546541279413094e-05, | |
| "loss": 0.3976, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.063100137174211, | |
| "grad_norm": 0.3291669273065511, | |
| "learning_rate": 5.799313504030545e-05, | |
| "loss": 0.3947, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.085048010973937, | |
| "grad_norm": 0.37474732162243923, | |
| "learning_rate": 5.743537673901485e-05, | |
| "loss": 0.3971, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.1069958847736627, | |
| "grad_norm": 0.46251544205958794, | |
| "learning_rate": 5.68734012824825e-05, | |
| "loss": 0.396, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.128943758573388, | |
| "grad_norm": 0.54656137847099, | |
| "learning_rate": 5.6307344597667555e-05, | |
| "loss": 0.4027, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.150891632373114, | |
| "grad_norm": 0.6049235984601056, | |
| "learning_rate": 5.5737343598670104e-05, | |
| "loss": 0.4005, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.5234566946423048, | |
| "learning_rate": 5.5163536153615185e-05, | |
| "loss": 0.3955, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.1947873799725652, | |
| "grad_norm": 0.4013877298974358, | |
| "learning_rate": 5.4586061051306204e-05, | |
| "loss": 0.4028, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.2167352537722906, | |
| "grad_norm": 0.3845465210881421, | |
| "learning_rate": 5.4005057967655634e-05, | |
| "loss": 0.3976, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.2386831275720165, | |
| "grad_norm": 0.408459585720323, | |
| "learning_rate": 5.3420667431901e-05, | |
| "loss": 0.3975, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.260631001371742, | |
| "grad_norm": 0.37864175367916847, | |
| "learning_rate": 5.283303079261471e-05, | |
| "loss": 0.3936, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.2825788751714677, | |
| "grad_norm": 0.3827028198880068, | |
| "learning_rate": 5.22422901835155e-05, | |
| "loss": 0.3936, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.3045267489711936, | |
| "grad_norm": 0.3484431146343949, | |
| "learning_rate": 5.164858848909009e-05, | |
| "loss": 0.388, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.326474622770919, | |
| "grad_norm": 0.3095716919680074, | |
| "learning_rate": 5.1052069310033216e-05, | |
| "loss": 0.3988, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.348422496570645, | |
| "grad_norm": 0.39008928135047216, | |
| "learning_rate": 5.0452876928514434e-05, | |
| "loss": 0.3966, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.44827732087715033, | |
| "learning_rate": 4.9851156273280064e-05, | |
| "loss": 0.3953, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.392318244170096, | |
| "grad_norm": 0.2796243136669204, | |
| "learning_rate": 4.92470528845988e-05, | |
| "loss": 0.3964, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.4142661179698215, | |
| "grad_norm": 0.3186398495286058, | |
| "learning_rate": 4.8640712879059354e-05, | |
| "loss": 0.3923, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.4362139917695473, | |
| "grad_norm": 0.35322375158512703, | |
| "learning_rate": 4.8032282914228743e-05, | |
| "loss": 0.3886, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.4581618655692727, | |
| "grad_norm": 0.18461936703017578, | |
| "learning_rate": 4.742191015317974e-05, | |
| "loss": 0.3917, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.4801097393689986, | |
| "grad_norm": 0.33683641542461873, | |
| "learning_rate": 4.680974222889595e-05, | |
| "loss": 0.394, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.5020576131687244, | |
| "grad_norm": 0.3169424542788605, | |
| "learning_rate": 4.6195927208563394e-05, | |
| "loss": 0.3871, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.52400548696845, | |
| "grad_norm": 0.15430100844009037, | |
| "learning_rate": 4.558061355775693e-05, | |
| "loss": 0.3857, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.5459533607681757, | |
| "grad_norm": 0.2820303722197663, | |
| "learning_rate": 4.496395010453038e-05, | |
| "loss": 0.3886, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.24770492735610264, | |
| "learning_rate": 4.4346086003418985e-05, | |
| "loss": 0.3882, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.589849108367627, | |
| "grad_norm": 0.15993713137779939, | |
| "learning_rate": 4.372717069936287e-05, | |
| "loss": 0.3891, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.611796982167353, | |
| "grad_norm": 0.24474244306163354, | |
| "learning_rate": 4.310735389156026e-05, | |
| "loss": 0.392, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.633744855967078, | |
| "grad_norm": 0.1905642453010369, | |
| "learning_rate": 4.248678549725923e-05, | |
| "loss": 0.3917, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.6556927297668036, | |
| "grad_norm": 0.15617551076050876, | |
| "learning_rate": 4.18656156154967e-05, | |
| "loss": 0.3945, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.6776406035665294, | |
| "grad_norm": 0.21098249998156268, | |
| "learning_rate": 4.124399449079348e-05, | |
| "loss": 0.3866, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.6995884773662553, | |
| "grad_norm": 0.17151651983910782, | |
| "learning_rate": 4.0622072476814045e-05, | |
| "loss": 0.3872, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.7215363511659807, | |
| "grad_norm": 0.14636727205415834, | |
| "learning_rate": 4e-05, | |
| "loss": 0.3919, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.7434842249657065, | |
| "grad_norm": 0.21883566171534793, | |
| "learning_rate": 3.937792752318597e-05, | |
| "loss": 0.3854, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.15504692956699287, | |
| "learning_rate": 3.8756005509206535e-05, | |
| "loss": 0.3883, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.787379972565158, | |
| "grad_norm": 0.21026818171396724, | |
| "learning_rate": 3.81343843845033e-05, | |
| "loss": 0.3883, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.8093278463648836, | |
| "grad_norm": 0.1846577189184653, | |
| "learning_rate": 3.751321450274078e-05, | |
| "loss": 0.3921, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.831275720164609, | |
| "grad_norm": 0.16874522256869426, | |
| "learning_rate": 3.689264610843975e-05, | |
| "loss": 0.3911, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.8532235939643344, | |
| "grad_norm": 0.16523585002563423, | |
| "learning_rate": 3.627282930063714e-05, | |
| "loss": 0.3886, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.8751714677640603, | |
| "grad_norm": 0.14959343517939727, | |
| "learning_rate": 3.565391399658102e-05, | |
| "loss": 0.392, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.897119341563786, | |
| "grad_norm": 0.13667256750591955, | |
| "learning_rate": 3.503604989546963e-05, | |
| "loss": 0.3846, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.9190672153635115, | |
| "grad_norm": 0.14202188950732816, | |
| "learning_rate": 3.4419386442243084e-05, | |
| "loss": 0.3902, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.9410150891632374, | |
| "grad_norm": 0.14821759155695716, | |
| "learning_rate": 3.380407279143661e-05, | |
| "loss": 0.3867, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.1667332312440918, | |
| "learning_rate": 3.3190257771104055e-05, | |
| "loss": 0.3927, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.9849108367626886, | |
| "grad_norm": 0.14452591810498722, | |
| "learning_rate": 3.257808984682027e-05, | |
| "loss": 0.3855, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 3.006858710562414, | |
| "grad_norm": 0.18251829909357556, | |
| "learning_rate": 3.196771708577127e-05, | |
| "loss": 0.3837, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 3.02880658436214, | |
| "grad_norm": 0.17364339659347747, | |
| "learning_rate": 3.135928712094067e-05, | |
| "loss": 0.359, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 3.0507544581618657, | |
| "grad_norm": 0.17747923559690468, | |
| "learning_rate": 3.075294711540123e-05, | |
| "loss": 0.3557, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 3.072702331961591, | |
| "grad_norm": 0.1787283180499878, | |
| "learning_rate": 3.0148843726719953e-05, | |
| "loss": 0.3516, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.094650205761317, | |
| "grad_norm": 0.19311413443289224, | |
| "learning_rate": 2.9547123071485586e-05, | |
| "loss": 0.3532, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 3.1165980795610424, | |
| "grad_norm": 0.15814186428435004, | |
| "learning_rate": 2.8947930689966798e-05, | |
| "loss": 0.3598, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 3.1385459533607682, | |
| "grad_norm": 0.17222178104775263, | |
| "learning_rate": 2.8351411510909926e-05, | |
| "loss": 0.3567, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.15304672551321433, | |
| "learning_rate": 2.7757709816484512e-05, | |
| "loss": 0.3634, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.1824417009602195, | |
| "grad_norm": 0.15457992995371023, | |
| "learning_rate": 2.71669692073853e-05, | |
| "loss": 0.3542, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.2043895747599453, | |
| "grad_norm": 0.1479490816574568, | |
| "learning_rate": 2.6579332568099006e-05, | |
| "loss": 0.3576, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 3.2263374485596708, | |
| "grad_norm": 0.15469338097636914, | |
| "learning_rate": 2.5994942032344376e-05, | |
| "loss": 0.357, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 3.2482853223593966, | |
| "grad_norm": 0.13891659760939676, | |
| "learning_rate": 2.54139389486938e-05, | |
| "loss": 0.3559, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 3.270233196159122, | |
| "grad_norm": 0.143615476750949, | |
| "learning_rate": 2.4836463846384832e-05, | |
| "loss": 0.3529, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 3.292181069958848, | |
| "grad_norm": 0.1420809451110692, | |
| "learning_rate": 2.4262656401329913e-05, | |
| "loss": 0.359, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.3141289437585733, | |
| "grad_norm": 0.14089610783501166, | |
| "learning_rate": 2.3692655402332455e-05, | |
| "loss": 0.3565, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 3.336076817558299, | |
| "grad_norm": 0.1402483457342098, | |
| "learning_rate": 2.3126598717517514e-05, | |
| "loss": 0.3508, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.12317925977087589, | |
| "learning_rate": 2.256462326098516e-05, | |
| "loss": 0.3558, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 3.3799725651577504, | |
| "grad_norm": 0.1452523632260013, | |
| "learning_rate": 2.200686495969457e-05, | |
| "loss": 0.3546, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 3.401920438957476, | |
| "grad_norm": 0.11788796867275445, | |
| "learning_rate": 2.1453458720586902e-05, | |
| "loss": 0.3562, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.4238683127572016, | |
| "grad_norm": 0.14156518711907154, | |
| "learning_rate": 2.0904538397954913e-05, | |
| "loss": 0.3595, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.4458161865569275, | |
| "grad_norm": 0.1152521891751902, | |
| "learning_rate": 2.0360236761067117e-05, | |
| "loss": 0.3572, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 3.467764060356653, | |
| "grad_norm": 0.1292120630664868, | |
| "learning_rate": 1.9820685462054413e-05, | |
| "loss": 0.3608, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 3.4897119341563787, | |
| "grad_norm": 0.12307673484565831, | |
| "learning_rate": 1.9286015004066984e-05, | |
| "loss": 0.3571, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 3.511659807956104, | |
| "grad_norm": 0.11387530947074216, | |
| "learning_rate": 1.8756354709708998e-05, | |
| "loss": 0.3593, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.53360768175583, | |
| "grad_norm": 0.1276131097306211, | |
| "learning_rate": 1.8231832689758903e-05, | |
| "loss": 0.3528, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.11562929678821103, | |
| "learning_rate": 1.771257581218287e-05, | |
| "loss": 0.3562, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 3.577503429355281, | |
| "grad_norm": 0.11195216588496365, | |
| "learning_rate": 1.7198709671448696e-05, | |
| "loss": 0.3568, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 3.599451303155007, | |
| "grad_norm": 0.10790427870554664, | |
| "learning_rate": 1.6690358558147967e-05, | |
| "loss": 0.3522, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 3.6213991769547325, | |
| "grad_norm": 0.11612616520204616, | |
| "learning_rate": 1.6187645428933372e-05, | |
| "loss": 0.3584, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.6433470507544583, | |
| "grad_norm": 0.11400211648602655, | |
| "learning_rate": 1.5690691876778746e-05, | |
| "loss": 0.3622, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 3.6652949245541837, | |
| "grad_norm": 0.1065206298660862, | |
| "learning_rate": 1.5199618101569003e-05, | |
| "loss": 0.3544, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 3.6872427983539096, | |
| "grad_norm": 0.10461159135503716, | |
| "learning_rate": 1.4714542881026947e-05, | |
| "loss": 0.3519, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.709190672153635, | |
| "grad_norm": 0.12285024075507715, | |
| "learning_rate": 1.4235583541984092e-05, | |
| "loss": 0.3621, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 3.731138545953361, | |
| "grad_norm": 0.10841173453177154, | |
| "learning_rate": 1.3762855932002404e-05, | |
| "loss": 0.3554, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.10185093830948505, | |
| "learning_rate": 1.3296474391353854e-05, | |
| "loss": 0.3602, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 3.775034293552812, | |
| "grad_norm": 0.10890679245723875, | |
| "learning_rate": 1.2836551725364572e-05, | |
| "loss": 0.3489, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.796982167352538, | |
| "grad_norm": 0.10191643511945396, | |
| "learning_rate": 1.2383199177130135e-05, | |
| "loss": 0.3536, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 3.8189300411522633, | |
| "grad_norm": 0.09752716587838449, | |
| "learning_rate": 1.1936526400608938e-05, | |
| "loss": 0.3517, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.840877914951989, | |
| "grad_norm": 0.09567252814321132, | |
| "learning_rate": 1.1496641434099725e-05, | |
| "loss": 0.3499, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.8628257887517146, | |
| "grad_norm": 0.10384979007917372, | |
| "learning_rate": 1.1063650674110011e-05, | |
| "loss": 0.3562, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.8847736625514404, | |
| "grad_norm": 0.09268596571493516, | |
| "learning_rate": 1.0637658849621593e-05, | |
| "loss": 0.3521, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.9067215363511663, | |
| "grad_norm": 0.08906274897668208, | |
| "learning_rate": 1.0218768996759399e-05, | |
| "loss": 0.3528, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.9286694101508917, | |
| "grad_norm": 0.09311271485149279, | |
| "learning_rate": 9.807082433869727e-06, | |
| "loss": 0.359, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.09966102704447338, | |
| "learning_rate": 9.402698737014098e-06, | |
| "loss": 0.3643, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.972565157750343, | |
| "grad_norm": 0.08428095680890077, | |
| "learning_rate": 9.005715715884409e-06, | |
| "loss": 0.3564, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.9945130315500688, | |
| "grad_norm": 0.098824142446022, | |
| "learning_rate": 8.616229390145361e-06, | |
| "loss": 0.3561, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 4.065843621399177, | |
| "grad_norm": 0.13982352922009822, | |
| "learning_rate": 8.23433396620986e-06, | |
| "loss": 0.341, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 4.0877914951989025, | |
| "grad_norm": 0.11357790917439833, | |
| "learning_rate": 7.86012181445297e-06, | |
| "loss": 0.3427, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 4.109739368998628, | |
| "grad_norm": 0.09508337442980856, | |
| "learning_rate": 7.4936834468699945e-06, | |
| "loss": 0.337, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 4.131687242798354, | |
| "grad_norm": 0.10345088268864072, | |
| "learning_rate": 7.135107495183975e-06, | |
| "loss": 0.3345, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 4.153635116598079, | |
| "grad_norm": 0.1048793029567903, | |
| "learning_rate": 6.784480689408099e-06, | |
| "loss": 0.3297, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 4.175582990397805, | |
| "grad_norm": 0.10481831928351991, | |
| "learning_rate": 6.441887836867962e-06, | |
| "loss": 0.3414, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 4.197530864197531, | |
| "grad_norm": 0.11340261438354692, | |
| "learning_rate": 6.107411801688905e-06, | |
| "loss": 0.3387, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 4.219478737997257, | |
| "grad_norm": 0.10558490986119158, | |
| "learning_rate": 5.781133484753451e-06, | |
| "loss": 0.3386, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.2414266117969825, | |
| "grad_norm": 0.09563394121242606, | |
| "learning_rate": 5.463131804133461e-06, | |
| "loss": 0.3392, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 4.2633744855967075, | |
| "grad_norm": 0.097884687531601, | |
| "learning_rate": 5.1534836760019824e-06, | |
| "loss": 0.3406, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 4.285322359396433, | |
| "grad_norm": 0.09586101042431322, | |
| "learning_rate": 4.852263996029259e-06, | |
| "loss": 0.3347, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 4.307270233196159, | |
| "grad_norm": 0.09449695957524798, | |
| "learning_rate": 4.559545621267414e-06, | |
| "loss": 0.3316, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 4.329218106995885, | |
| "grad_norm": 0.09218455432645911, | |
| "learning_rate": 4.275399352528342e-06, | |
| "loss": 0.3382, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 4.35116598079561, | |
| "grad_norm": 0.09122205514695, | |
| "learning_rate": 3.999893917258799e-06, | |
| "loss": 0.3421, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 4.373113854595336, | |
| "grad_norm": 0.08884394525292823, | |
| "learning_rate": 3.733095952917101e-06, | |
| "loss": 0.3359, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 4.395061728395062, | |
| "grad_norm": 0.08788073543014369, | |
| "learning_rate": 3.4750699908552464e-06, | |
| "loss": 0.3348, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 4.4170096021947876, | |
| "grad_norm": 0.0851504561450446, | |
| "learning_rate": 3.225878440710544e-06, | |
| "loss": 0.338, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 4.438957475994513, | |
| "grad_norm": 0.08806355910065146, | |
| "learning_rate": 2.9855815753103436e-06, | |
| "loss": 0.3408, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.460905349794238, | |
| "grad_norm": 0.08443831027497156, | |
| "learning_rate": 2.754237516093623e-06, | |
| "loss": 0.3336, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 4.482853223593964, | |
| "grad_norm": 0.0846796048507417, | |
| "learning_rate": 2.5319022190529997e-06, | |
| "loss": 0.3347, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 4.50480109739369, | |
| "grad_norm": 0.08433609787768302, | |
| "learning_rate": 2.3186294612004365e-06, | |
| "loss": 0.3359, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 4.526748971193416, | |
| "grad_norm": 0.08573881642215767, | |
| "learning_rate": 2.1144708275599955e-06, | |
| "loss": 0.3411, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 4.548696844993142, | |
| "grad_norm": 0.08265953354626032, | |
| "learning_rate": 1.9194756986908025e-06, | |
| "loss": 0.3351, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 4.570644718792867, | |
| "grad_norm": 0.07968266677719599, | |
| "learning_rate": 1.7336912387432115e-06, | |
| "loss": 0.3329, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 4.592592592592593, | |
| "grad_norm": 0.07672083627731244, | |
| "learning_rate": 1.5571623840510185e-06, | |
| "loss": 0.3344, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 4.614540466392318, | |
| "grad_norm": 0.07767236853772148, | |
| "learning_rate": 1.3899318322625744e-06, | |
| "loss": 0.3353, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 4.636488340192044, | |
| "grad_norm": 0.079617742448113, | |
| "learning_rate": 1.2320400320133551e-06, | |
| "loss": 0.3351, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 4.658436213991769, | |
| "grad_norm": 0.07817205047145995, | |
| "learning_rate": 1.0835251731425013e-06, | |
| "loss": 0.3389, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.680384087791495, | |
| "grad_norm": 0.079996413423676, | |
| "learning_rate": 9.444231774557199e-07, | |
| "loss": 0.3407, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 4.702331961591221, | |
| "grad_norm": 0.07582115660185369, | |
| "learning_rate": 8.147676900367308e-07, | |
| "loss": 0.3412, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 4.724279835390947, | |
| "grad_norm": 0.07655208386535386, | |
| "learning_rate": 6.945900711094534e-07, | |
| "loss": 0.3308, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 4.746227709190672, | |
| "grad_norm": 0.07575601990026438, | |
| "learning_rate": 5.839193884527472e-07, | |
| "loss": 0.3326, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 4.768175582990398, | |
| "grad_norm": 0.07445403870736927, | |
| "learning_rate": 4.827824103697332e-07, | |
| "loss": 0.3312, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.790123456790123, | |
| "grad_norm": 0.07565335743085141, | |
| "learning_rate": 3.912035992132257e-07, | |
| "loss": 0.3396, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 4.812071330589849, | |
| "grad_norm": 0.07504738171668904, | |
| "learning_rate": 3.0920510546894156e-07, | |
| "loss": 0.3279, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 4.834019204389575, | |
| "grad_norm": 0.07495711123910379, | |
| "learning_rate": 2.3680676239789647e-07, | |
| "loss": 0.3373, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 4.8559670781893, | |
| "grad_norm": 0.07581516860262182, | |
| "learning_rate": 1.740260812392558e-07, | |
| "loss": 0.3377, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 4.877914951989026, | |
| "grad_norm": 0.07440451827362232, | |
| "learning_rate": 1.208782469748293e-07, | |
| "loss": 0.3375, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.899862825788752, | |
| "grad_norm": 0.07405892919582897, | |
| "learning_rate": 7.737611465622686e-08, | |
| "loss": 0.3314, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 4.921810699588478, | |
| "grad_norm": 0.07208525046101905, | |
| "learning_rate": 4.353020629556781e-08, | |
| "loss": 0.334, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 4.9437585733882035, | |
| "grad_norm": 0.07250510923845269, | |
| "learning_rate": 1.934870832047686e-08, | |
| "loss": 0.3358, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 4.965706447187928, | |
| "grad_norm": 0.07367572989174814, | |
| "learning_rate": 4.837469594018984e-09, | |
| "loss": 0.3436, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.987654320987654, | |
| "grad_norm": 0.07408156451999265, | |
| "learning_rate": 0.0, | |
| "loss": 0.337, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.987654320987654, | |
| "step": 225, | |
| "total_flos": 5.796779967165497e+18, | |
| "train_loss": 0.06433244016435412, | |
| "train_runtime": 4471.4422, | |
| "train_samples_per_second": 26.061, | |
| "train_steps_per_second": 0.05 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 225, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.796779967165497e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |