| { | |
| "best_metric": 0.10067134350538254, | |
| "best_model_checkpoint": "saves/psy-course-doc/Llama-3.1-8B-Instruct/train/fold7/checkpoint-80", | |
| "epoch": 4.938271604938271, | |
| "eval_steps": 10, | |
| "global_step": 125, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03950617283950617, | |
| "grad_norm": 0.17030450701713562, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.2112, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 0.17128004133701324, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.2165, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 0.18140414357185364, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.1579, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 0.17117320001125336, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.1965, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.17991754412651062, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.1985, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 0.17068563401699066, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.2192, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 0.1799967885017395, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.2599, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 0.1989358812570572, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.2001, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 0.1522873342037201, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.1285, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 0.151645228266716, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.1825, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.18109993636608124, | |
| "eval_runtime": 22.2726, | |
| "eval_samples_per_second": 2.02, | |
| "eval_steps_per_second": 2.02, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 0.15675899386405945, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.115, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 0.17178775370121002, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.1915, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.14882557094097137, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0953, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.1637958288192749, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.2168, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.151862233877182, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.1155, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.15528938174247742, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.0825, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.14563225209712982, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.1099, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.1248278096318245, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.1319, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.11705093830823898, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.1604, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.10367219150066376, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.0806, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.12763603031635284, | |
| "eval_runtime": 22.2738, | |
| "eval_samples_per_second": 2.02, | |
| "eval_steps_per_second": 2.02, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.10195054858922958, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.0846, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.08751470595598221, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.0704, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.09555941075086594, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.1247, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.10090352594852448, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.1123, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.10310988128185272, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.1255, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.18661224842071533, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.1918, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.11039568483829498, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.104, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.09591447561979294, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.0857, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.10397636890411377, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.0678, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.11892275512218475, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.111, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.11317049711942673, | |
| "eval_runtime": 22.2771, | |
| "eval_samples_per_second": 2.02, | |
| "eval_steps_per_second": 2.02, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.10336616635322571, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.0828, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.11497625708580017, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.1268, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.10329342633485794, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.0508, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.09853846579790115, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.0388, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.10033365339040756, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1177, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.1055055782198906, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.0767, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.10473234206438065, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1166, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.1142139807343483, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.0572, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.1204463541507721, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.0853, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.09684475511312485, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.0667, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.10802517831325531, | |
| "eval_runtime": 22.2891, | |
| "eval_samples_per_second": 2.019, | |
| "eval_steps_per_second": 2.019, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.10067864507436752, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.0753, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.09055954217910767, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.0686, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.08414781838655472, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.0756, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.09058612585067749, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.0976, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.10440947860479355, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.0717, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.12771636247634888, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.0707, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.10225579887628555, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1444, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.07501435279846191, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.0646, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.10964561998844147, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1034, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.10926580429077148, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.098, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.10522958636283875, | |
| "eval_runtime": 22.316, | |
| "eval_samples_per_second": 2.016, | |
| "eval_steps_per_second": 2.016, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.14977577328681946, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.0843, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.08431752771139145, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.0893, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.06914456188678741, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.0567, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.07669289410114288, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.0875, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.1036219522356987, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1455, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.07544882595539093, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.0914, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.10492753237485886, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1377, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.10346435755491257, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.1066, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.07775168865919113, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.0598, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.09240282326936722, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.0343, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.10299863666296005, | |
| "eval_runtime": 22.288, | |
| "eval_samples_per_second": 2.019, | |
| "eval_steps_per_second": 2.019, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.10413602739572525, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.0956, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.08183609694242477, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.0588, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.07852060347795486, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.055, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.09505973011255264, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.0689, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.12405043840408325, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.1367, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.11045446991920471, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.0622, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.10130424797534943, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.0762, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.07693851739168167, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.0457, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.08287542313337326, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0156, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.10197284817695618, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.0541, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.10247812420129776, | |
| "eval_runtime": 22.2826, | |
| "eval_samples_per_second": 2.02, | |
| "eval_steps_per_second": 2.02, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.1042519360780716, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.0536, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.07915712893009186, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.0411, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.08079107850790024, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.0631, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.08635604381561279, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.0491, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.09656098484992981, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.0644, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 0.14779677987098694, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1134, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.10362217575311661, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.1143, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.08608033508062363, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.08, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.07910611480474472, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0512, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.09638766944408417, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.1154, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.10067134350538254, | |
| "eval_runtime": 22.2998, | |
| "eval_samples_per_second": 2.018, | |
| "eval_steps_per_second": 2.018, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.09143459796905518, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0757, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.09514543414115906, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0786, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.08326562494039536, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.087, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.09207258373498917, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.0942, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.07398128509521484, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0452, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.08139006048440933, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.0468, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.08422186970710754, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.0662, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.09458501636981964, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.0761, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.09712503105401993, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0678, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.09837882965803146, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0901, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.10096059739589691, | |
| "eval_runtime": 22.3213, | |
| "eval_samples_per_second": 2.016, | |
| "eval_steps_per_second": 2.016, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.08770782500505447, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.0654, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.08923909813165665, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.0697, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.08821766078472137, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0263, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.09415401518344879, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.0429, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.0986524149775505, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.0398, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.08546709269285202, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0139, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.08082624524831772, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0449, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.09882638603448868, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.0829, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.10103920102119446, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.0709, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.09675954282283783, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0529, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.1010698601603508, | |
| "eval_runtime": 22.3145, | |
| "eval_samples_per_second": 2.017, | |
| "eval_steps_per_second": 2.017, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.14186929166316986, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.0991, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.13005265593528748, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.0653, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.09086073935031891, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0755, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.10105685144662857, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0912, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.10067118704319, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0812, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.08738145232200623, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0547, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.10857798904180527, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0898, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.0749911516904831, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.0355, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.10380955785512924, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.0941, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.08231041580438614, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0544, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.10101938247680664, | |
| "eval_runtime": 22.329, | |
| "eval_samples_per_second": 2.015, | |
| "eval_steps_per_second": 2.015, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.07592027634382248, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0327, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.07671812921762466, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.044, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.10160104930400848, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0464, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.10133630782365799, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0746, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.10026230663061142, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0572, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.08998555690050125, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.0789, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.11149641871452332, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0975, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.08201292157173157, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.039, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.09800563007593155, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.0713, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.11083582043647766, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.0964, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.10088521987199783, | |
| "eval_runtime": 22.3022, | |
| "eval_samples_per_second": 2.018, | |
| "eval_steps_per_second": 2.018, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.0909806415438652, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.0643, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.07180990278720856, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0287, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.08601696044206619, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.0572, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.08180883526802063, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.0524, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.09118040651082993, | |
| "learning_rate": 0.0, | |
| "loss": 0.0694, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.506946920897577e+17, | |
| "train_loss": 0.08986351612955332, | |
| "train_runtime": 3744.6347, | |
| "train_samples_per_second": 0.541, | |
| "train_steps_per_second": 0.033 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 125, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.506946920897577e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |