| { | |
| "best_metric": 0.32943063974380493, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-350", | |
| "epoch": 2.908713692946058, | |
| "eval_steps": 50, | |
| "global_step": 350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008298755186721992, | |
| "grad_norm": 11.30918025970459, | |
| "learning_rate": 1e-05, | |
| "loss": 4.6137, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008298755186721992, | |
| "eval_loss": 5.352054595947266, | |
| "eval_runtime": 7.0594, | |
| "eval_samples_per_second": 28.756, | |
| "eval_steps_per_second": 7.224, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.016597510373443983, | |
| "grad_norm": 13.656156539916992, | |
| "learning_rate": 2e-05, | |
| "loss": 4.4637, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.024896265560165973, | |
| "grad_norm": 12.059760093688965, | |
| "learning_rate": 3e-05, | |
| "loss": 4.3758, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03319502074688797, | |
| "grad_norm": 10.070697784423828, | |
| "learning_rate": 4e-05, | |
| "loss": 4.6825, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.04149377593360996, | |
| "grad_norm": 10.277752876281738, | |
| "learning_rate": 5e-05, | |
| "loss": 4.4531, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04979253112033195, | |
| "grad_norm": 8.195782661437988, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9268, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.058091286307053944, | |
| "grad_norm": 9.387673377990723, | |
| "learning_rate": 7e-05, | |
| "loss": 3.7347, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06639004149377593, | |
| "grad_norm": 8.870927810668945, | |
| "learning_rate": 8e-05, | |
| "loss": 3.4846, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07468879668049792, | |
| "grad_norm": 6.948357582092285, | |
| "learning_rate": 9e-05, | |
| "loss": 3.3592, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08298755186721991, | |
| "grad_norm": 5.063608169555664, | |
| "learning_rate": 0.0001, | |
| "loss": 2.8956, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0912863070539419, | |
| "grad_norm": 14.618287086486816, | |
| "learning_rate": 9.999799726899262e-05, | |
| "loss": 2.92, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0995850622406639, | |
| "grad_norm": 18.60139274597168, | |
| "learning_rate": 9.999198923640774e-05, | |
| "loss": 3.0395, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1078838174273859, | |
| "grad_norm": 16.003385543823242, | |
| "learning_rate": 9.998197638354428e-05, | |
| "loss": 2.8771, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.11618257261410789, | |
| "grad_norm": 7.36771297454834, | |
| "learning_rate": 9.996795951252427e-05, | |
| "loss": 2.5778, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.12448132780082988, | |
| "grad_norm": 7.226404190063477, | |
| "learning_rate": 9.994993974622862e-05, | |
| "loss": 2.9167, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.13278008298755187, | |
| "grad_norm": 9.766325950622559, | |
| "learning_rate": 9.992791852820709e-05, | |
| "loss": 2.9431, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.14107883817427386, | |
| "grad_norm": 5.852901935577393, | |
| "learning_rate": 9.990189762256276e-05, | |
| "loss": 2.3363, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.14937759336099585, | |
| "grad_norm": 5.548067569732666, | |
| "learning_rate": 9.987187911381059e-05, | |
| "loss": 2.237, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.15767634854771784, | |
| "grad_norm": 6.030773639678955, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 2.1454, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.16597510373443983, | |
| "grad_norm": 5.069052696228027, | |
| "learning_rate": 9.979985922607475e-05, | |
| "loss": 1.971, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.17427385892116182, | |
| "grad_norm": 5.834778308868408, | |
| "learning_rate": 9.97578636165496e-05, | |
| "loss": 2.5553, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1825726141078838, | |
| "grad_norm": 4.660792350769043, | |
| "learning_rate": 9.97118819423714e-05, | |
| "loss": 2.2076, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1908713692946058, | |
| "grad_norm": 4.532413959503174, | |
| "learning_rate": 9.966191788709716e-05, | |
| "loss": 2.1984, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1991701244813278, | |
| "grad_norm": 6.078835964202881, | |
| "learning_rate": 9.960797545330936e-05, | |
| "loss": 2.5873, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2074688796680498, | |
| "grad_norm": 5.118812084197998, | |
| "learning_rate": 9.955005896229543e-05, | |
| "loss": 2.7898, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2157676348547718, | |
| "grad_norm": 5.2867231369018555, | |
| "learning_rate": 9.948817305370143e-05, | |
| "loss": 2.2812, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.22406639004149378, | |
| "grad_norm": 4.749370098114014, | |
| "learning_rate": 9.94223226851605e-05, | |
| "loss": 2.4715, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.23236514522821577, | |
| "grad_norm": 4.7184977531433105, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 2.3676, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.24066390041493776, | |
| "grad_norm": 5.203877925872803, | |
| "learning_rate": 9.927874998629714e-05, | |
| "loss": 2.4597, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.24896265560165975, | |
| "grad_norm": 7.009748935699463, | |
| "learning_rate": 9.92010391574745e-05, | |
| "loss": 2.7419, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2572614107883817, | |
| "grad_norm": 4.794604301452637, | |
| "learning_rate": 9.911938687078324e-05, | |
| "loss": 2.4883, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.26556016597510373, | |
| "grad_norm": 4.864479064941406, | |
| "learning_rate": 9.9033799667326e-05, | |
| "loss": 2.2448, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.27385892116182575, | |
| "grad_norm": 4.177216529846191, | |
| "learning_rate": 9.89442844034286e-05, | |
| "loss": 2.1125, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2821576763485477, | |
| "grad_norm": 3.661107063293457, | |
| "learning_rate": 9.885084825009086e-05, | |
| "loss": 1.8155, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.29045643153526973, | |
| "grad_norm": 4.322318077087402, | |
| "learning_rate": 9.875349869241201e-05, | |
| "loss": 2.1389, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2987551867219917, | |
| "grad_norm": 4.690096378326416, | |
| "learning_rate": 9.865224352899119e-05, | |
| "loss": 2.0913, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3070539419087137, | |
| "grad_norm": 4.660186767578125, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 2.1136, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3153526970954357, | |
| "grad_norm": 4.324281215667725, | |
| "learning_rate": 9.843804914304577e-05, | |
| "loss": 2.2243, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3236514522821577, | |
| "grad_norm": 4.355935096740723, | |
| "learning_rate": 9.83251270794707e-05, | |
| "loss": 2.0655, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.33195020746887965, | |
| "grad_norm": 4.537855625152588, | |
| "learning_rate": 9.820833372667812e-05, | |
| "loss": 2.4219, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.34024896265560167, | |
| "grad_norm": 3.879166603088379, | |
| "learning_rate": 9.80876784408948e-05, | |
| "loss": 2.0775, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.34854771784232363, | |
| "grad_norm": 4.542738437652588, | |
| "learning_rate": 9.796317088772403e-05, | |
| "loss": 2.3691, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.35684647302904565, | |
| "grad_norm": 3.76499342918396, | |
| "learning_rate": 9.783482104137127e-05, | |
| "loss": 1.6421, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3651452282157676, | |
| "grad_norm": 4.638901233673096, | |
| "learning_rate": 9.770263918384524e-05, | |
| "loss": 2.1609, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.37344398340248963, | |
| "grad_norm": 4.964886665344238, | |
| "learning_rate": 9.75666359041341e-05, | |
| "loss": 1.5742, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3817427385892116, | |
| "grad_norm": 4.453831672668457, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 2.2456, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3900414937759336, | |
| "grad_norm": 5.158775806427002, | |
| "learning_rate": 9.728320896389263e-05, | |
| "loss": 2.1847, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3983402489626556, | |
| "grad_norm": 5.11721134185791, | |
| "learning_rate": 9.713580800847916e-05, | |
| "loss": 1.7405, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4066390041493776, | |
| "grad_norm": 5.306658744812012, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.6412, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.4149377593360996, | |
| "grad_norm": 5.626243591308594, | |
| "learning_rate": 9.682969016701358e-05, | |
| "loss": 1.9538, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4149377593360996, | |
| "eval_loss": 1.9951374530792236, | |
| "eval_runtime": 7.1371, | |
| "eval_samples_per_second": 28.443, | |
| "eval_steps_per_second": 7.146, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.42323651452282157, | |
| "grad_norm": 5.736331462860107, | |
| "learning_rate": 9.66709978038292e-05, | |
| "loss": 2.2465, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4315352697095436, | |
| "grad_norm": 5.308639049530029, | |
| "learning_rate": 9.650856666246693e-05, | |
| "loss": 2.1472, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.43983402489626555, | |
| "grad_norm": 4.070306301116943, | |
| "learning_rate": 9.634240975516209e-05, | |
| "loss": 1.5106, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.44813278008298757, | |
| "grad_norm": 4.719001293182373, | |
| "learning_rate": 9.617254039261834e-05, | |
| "loss": 1.5839, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.45643153526970953, | |
| "grad_norm": 8.284409523010254, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 1.9048, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.46473029045643155, | |
| "grad_norm": 5.499932765960693, | |
| "learning_rate": 9.582171903054816e-05, | |
| "loss": 2.0164, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.4730290456431535, | |
| "grad_norm": 4.232677936553955, | |
| "learning_rate": 9.564079513505455e-05, | |
| "loss": 1.6968, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.48132780082987553, | |
| "grad_norm": 6.6102776527404785, | |
| "learning_rate": 9.54562149901362e-05, | |
| "loss": 2.1707, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4896265560165975, | |
| "grad_norm": 5.499170303344727, | |
| "learning_rate": 9.526799338236827e-05, | |
| "loss": 2.1749, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.4979253112033195, | |
| "grad_norm": 6.410165309906006, | |
| "learning_rate": 9.507614539004082e-05, | |
| "loss": 2.0269, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5062240663900415, | |
| "grad_norm": 4.395339488983154, | |
| "learning_rate": 9.48806863819507e-05, | |
| "loss": 1.5799, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5145228215767634, | |
| "grad_norm": 5.214993953704834, | |
| "learning_rate": 9.468163201617062e-05, | |
| "loss": 1.4865, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5228215767634855, | |
| "grad_norm": 7.040699005126953, | |
| "learning_rate": 9.447899823879456e-05, | |
| "loss": 2.0153, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5311203319502075, | |
| "grad_norm": 5.472765922546387, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 1.9123, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5394190871369294, | |
| "grad_norm": 5.13813591003418, | |
| "learning_rate": 9.406305766604995e-05, | |
| "loss": 1.8829, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5477178423236515, | |
| "grad_norm": 5.717247486114502, | |
| "learning_rate": 9.384978419136468e-05, | |
| "loss": 2.0948, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5560165975103735, | |
| "grad_norm": 5.990386962890625, | |
| "learning_rate": 9.363299794378073e-05, | |
| "loss": 1.8932, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5643153526970954, | |
| "grad_norm": 7.660478591918945, | |
| "learning_rate": 9.341271628987968e-05, | |
| "loss": 2.3816, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5726141078838174, | |
| "grad_norm": 4.8774027824401855, | |
| "learning_rate": 9.318895687625751e-05, | |
| "loss": 1.6324, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5809128630705395, | |
| "grad_norm": 5.0475568771362305, | |
| "learning_rate": 9.296173762811085e-05, | |
| "loss": 1.5918, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5892116182572614, | |
| "grad_norm": 4.739609718322754, | |
| "learning_rate": 9.273107674780102e-05, | |
| "loss": 1.3994, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5975103734439834, | |
| "grad_norm": 6.396079063415527, | |
| "learning_rate": 9.249699271339593e-05, | |
| "loss": 1.8884, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6058091286307054, | |
| "grad_norm": 6.702268600463867, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 2.0607, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6141078838174274, | |
| "grad_norm": 4.683436393737793, | |
| "learning_rate": 9.201863046420065e-05, | |
| "loss": 1.458, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6224066390041494, | |
| "grad_norm": 5.798123359680176, | |
| "learning_rate": 9.177439057064683e-05, | |
| "loss": 1.9928, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6307053941908713, | |
| "grad_norm": 6.20752477645874, | |
| "learning_rate": 9.152680416240059e-05, | |
| "loss": 2.0453, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6390041493775933, | |
| "grad_norm": 4.803196430206299, | |
| "learning_rate": 9.1275891073421e-05, | |
| "loss": 1.6836, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6473029045643154, | |
| "grad_norm": 5.217916011810303, | |
| "learning_rate": 9.102167140416503e-05, | |
| "loss": 1.5624, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6556016597510373, | |
| "grad_norm": 5.689030170440674, | |
| "learning_rate": 9.076416551997721e-05, | |
| "loss": 1.2365, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6639004149377593, | |
| "grad_norm": 5.20029878616333, | |
| "learning_rate": 9.050339404945833e-05, | |
| "loss": 1.5095, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6721991701244814, | |
| "grad_norm": 7.432442665100098, | |
| "learning_rate": 9.023937788281278e-05, | |
| "loss": 2.1264, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.6804979253112033, | |
| "grad_norm": 5.681693077087402, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 1.5059, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6887966804979253, | |
| "grad_norm": 6.367238521575928, | |
| "learning_rate": 8.970169631991556e-05, | |
| "loss": 1.5183, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6970954356846473, | |
| "grad_norm": 7.021786212921143, | |
| "learning_rate": 8.942807399692543e-05, | |
| "loss": 1.8657, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7053941908713693, | |
| "grad_norm": 6.1124587059021, | |
| "learning_rate": 8.915129312088112e-05, | |
| "loss": 1.7342, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7136929460580913, | |
| "grad_norm": 6.746445655822754, | |
| "learning_rate": 8.88713758644883e-05, | |
| "loss": 1.7254, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7219917012448133, | |
| "grad_norm": 5.889866828918457, | |
| "learning_rate": 8.858834465170576e-05, | |
| "loss": 1.6482, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7302904564315352, | |
| "grad_norm": 6.273972511291504, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 1.5594, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7385892116182573, | |
| "grad_norm": 8.849343299865723, | |
| "learning_rate": 8.801303129827352e-05, | |
| "loss": 1.4801, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7468879668049793, | |
| "grad_norm": 6.7875566482543945, | |
| "learning_rate": 8.772079524553951e-05, | |
| "loss": 1.6455, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7551867219917012, | |
| "grad_norm": 5.853145599365234, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 1.3349, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.7634854771784232, | |
| "grad_norm": 7.308794021606445, | |
| "learning_rate": 8.712728144020118e-05, | |
| "loss": 1.8087, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7717842323651453, | |
| "grad_norm": 7.031892776489258, | |
| "learning_rate": 8.682605123353685e-05, | |
| "loss": 1.3419, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.7800829875518672, | |
| "grad_norm": 6.652479648590088, | |
| "learning_rate": 8.652187091988517e-05, | |
| "loss": 1.645, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.7883817427385892, | |
| "grad_norm": 8.24259090423584, | |
| "learning_rate": 8.621476486689992e-05, | |
| "loss": 1.6243, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.7966804979253111, | |
| "grad_norm": 5.5381975173950195, | |
| "learning_rate": 8.59047576766137e-05, | |
| "loss": 1.282, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8049792531120332, | |
| "grad_norm": 6.3423309326171875, | |
| "learning_rate": 8.559187418346703e-05, | |
| "loss": 1.3701, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8132780082987552, | |
| "grad_norm": 6.963042736053467, | |
| "learning_rate": 8.527613945231885e-05, | |
| "loss": 1.1813, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8215767634854771, | |
| "grad_norm": 8.486031532287598, | |
| "learning_rate": 8.495757877643858e-05, | |
| "loss": 1.5802, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8298755186721992, | |
| "grad_norm": 8.076821327209473, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 1.6718, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8298755186721992, | |
| "eval_loss": 1.3937488794326782, | |
| "eval_runtime": 7.2083, | |
| "eval_samples_per_second": 28.162, | |
| "eval_steps_per_second": 7.075, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8381742738589212, | |
| "grad_norm": 4.582700252532959, | |
| "learning_rate": 8.43120818934367e-05, | |
| "loss": 1.2412, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8464730290456431, | |
| "grad_norm": 11.859997749328613, | |
| "learning_rate": 8.398519739657996e-05, | |
| "loss": 1.9913, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.8547717842323651, | |
| "grad_norm": 7.4468841552734375, | |
| "learning_rate": 8.365559037137852e-05, | |
| "loss": 1.9243, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8630705394190872, | |
| "grad_norm": 7.206920146942139, | |
| "learning_rate": 8.332328722240073e-05, | |
| "loss": 1.7109, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8713692946058091, | |
| "grad_norm": 6.83170223236084, | |
| "learning_rate": 8.298831457019942e-05, | |
| "loss": 1.3447, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.8796680497925311, | |
| "grad_norm": 7.21415376663208, | |
| "learning_rate": 8.265069924917925e-05, | |
| "loss": 1.4393, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.8879668049792531, | |
| "grad_norm": 9.88570785522461, | |
| "learning_rate": 8.231046830544716e-05, | |
| "loss": 2.0253, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.8962655601659751, | |
| "grad_norm": 6.444344520568848, | |
| "learning_rate": 8.196764899464551e-05, | |
| "loss": 1.3066, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9045643153526971, | |
| "grad_norm": 5.2003326416015625, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 0.9973, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9128630705394191, | |
| "grad_norm": 5.82247257232666, | |
| "learning_rate": 8.127435532896388e-05, | |
| "loss": 1.2795, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.921161825726141, | |
| "grad_norm": 8.667819023132324, | |
| "learning_rate": 8.092393651331275e-05, | |
| "loss": 1.9045, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9294605809128631, | |
| "grad_norm": 7.90700626373291, | |
| "learning_rate": 8.057104040460062e-05, | |
| "loss": 1.3316, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9377593360995851, | |
| "grad_norm": 5.512808322906494, | |
| "learning_rate": 8.021569527306662e-05, | |
| "loss": 1.0711, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.946058091286307, | |
| "grad_norm": 8.198836326599121, | |
| "learning_rate": 7.985792958513931e-05, | |
| "loss": 1.7327, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.9543568464730291, | |
| "grad_norm": 9.416217803955078, | |
| "learning_rate": 7.949777200115616e-05, | |
| "loss": 1.6381, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9626556016597511, | |
| "grad_norm": 7.902495384216309, | |
| "learning_rate": 7.913525137306756e-05, | |
| "loss": 1.4719, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.970954356846473, | |
| "grad_norm": 6.16163444519043, | |
| "learning_rate": 7.877039674212569e-05, | |
| "loss": 1.1846, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.979253112033195, | |
| "grad_norm": 7.932492256164551, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 1.3338, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.9875518672199171, | |
| "grad_norm": 6.864604949951172, | |
| "learning_rate": 7.803380256922495e-05, | |
| "loss": 1.2608, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.995850622406639, | |
| "grad_norm": 8.558213233947754, | |
| "learning_rate": 7.76621220352657e-05, | |
| "loss": 1.271, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0062240663900415, | |
| "grad_norm": 14.87498664855957, | |
| "learning_rate": 7.728822550972523e-05, | |
| "loss": 1.9047, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0145228215767634, | |
| "grad_norm": 4.546281337738037, | |
| "learning_rate": 7.69121429451702e-05, | |
| "loss": 0.6296, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0228215767634854, | |
| "grad_norm": 5.022426128387451, | |
| "learning_rate": 7.653390446928909e-05, | |
| "loss": 0.7982, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.0311203319502074, | |
| "grad_norm": 6.32510232925415, | |
| "learning_rate": 7.615354038247888e-05, | |
| "loss": 0.895, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.0394190871369295, | |
| "grad_norm": 5.873676300048828, | |
| "learning_rate": 7.577108115541761e-05, | |
| "loss": 0.7959, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0477178423236515, | |
| "grad_norm": 6.671182632446289, | |
| "learning_rate": 7.53865574266234e-05, | |
| "loss": 0.7174, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.0560165975103735, | |
| "grad_norm": 5.916552543640137, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.516, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.0643153526970954, | |
| "grad_norm": 5.885247230529785, | |
| "learning_rate": 7.461143984236924e-05, | |
| "loss": 0.5581, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.0726141078838174, | |
| "grad_norm": 5.941064834594727, | |
| "learning_rate": 7.422090808099014e-05, | |
| "loss": 0.5006, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.0809128630705394, | |
| "grad_norm": 6.601113319396973, | |
| "learning_rate": 7.38284360010654e-05, | |
| "loss": 0.8121, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0892116182572613, | |
| "grad_norm": 6.470757484436035, | |
| "learning_rate": 7.343405504323519e-05, | |
| "loss": 0.7346, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.0975103734439835, | |
| "grad_norm": 5.690014839172363, | |
| "learning_rate": 7.303779680105843e-05, | |
| "loss": 0.5846, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.1058091286307055, | |
| "grad_norm": 7.497930526733398, | |
| "learning_rate": 7.263969301848188e-05, | |
| "loss": 1.1355, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1141078838174274, | |
| "grad_norm": 5.633259296417236, | |
| "learning_rate": 7.223977558729706e-05, | |
| "loss": 0.4784, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1224066390041494, | |
| "grad_norm": 5.755221843719482, | |
| "learning_rate": 7.183807654458564e-05, | |
| "loss": 0.6109, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1307053941908713, | |
| "grad_norm": 5.960432052612305, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 0.7443, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.1390041493775933, | |
| "grad_norm": 6.3896403312683105, | |
| "learning_rate": 7.102946248394909e-05, | |
| "loss": 0.7038, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.1473029045643153, | |
| "grad_norm": 4.847652435302734, | |
| "learning_rate": 7.062261224348203e-05, | |
| "loss": 0.4339, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.1556016597510372, | |
| "grad_norm": 3.1685574054718018, | |
| "learning_rate": 7.021410994121525e-05, | |
| "loss": 0.2151, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.1639004149377594, | |
| "grad_norm": 4.961888790130615, | |
| "learning_rate": 6.980398830195785e-05, | |
| "loss": 0.4, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1721991701244814, | |
| "grad_norm": 5.239101886749268, | |
| "learning_rate": 6.939228018024275e-05, | |
| "loss": 0.6228, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.1804979253112033, | |
| "grad_norm": 5.622098922729492, | |
| "learning_rate": 6.897901855769483e-05, | |
| "loss": 0.3817, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.1887966804979253, | |
| "grad_norm": 4.465243816375732, | |
| "learning_rate": 6.856423654038868e-05, | |
| "loss": 0.4005, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.1970954356846473, | |
| "grad_norm": 6.165435791015625, | |
| "learning_rate": 6.814796735619663e-05, | |
| "loss": 0.4155, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.2053941908713692, | |
| "grad_norm": 8.859424591064453, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 0.8074, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2136929460580912, | |
| "grad_norm": 6.36492919921875, | |
| "learning_rate": 6.731110099165164e-05, | |
| "loss": 0.4823, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2219917012448134, | |
| "grad_norm": 5.6012678146362305, | |
| "learning_rate": 6.689057085202737e-05, | |
| "loss": 0.4968, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.2302904564315353, | |
| "grad_norm": 7.916935920715332, | |
| "learning_rate": 6.646868762160399e-05, | |
| "loss": 0.5695, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.2385892116182573, | |
| "grad_norm": 5.010058403015137, | |
| "learning_rate": 6.604548509712658e-05, | |
| "loss": 0.3488, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.2468879668049793, | |
| "grad_norm": 9.163707733154297, | |
| "learning_rate": 6.562099718102788e-05, | |
| "loss": 0.6332, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2468879668049793, | |
| "eval_loss": 0.9936810731887817, | |
| "eval_runtime": 7.1349, | |
| "eval_samples_per_second": 28.452, | |
| "eval_steps_per_second": 7.148, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2551867219917012, | |
| "grad_norm": 7.536319255828857, | |
| "learning_rate": 6.519525787871235e-05, | |
| "loss": 0.6467, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.2634854771784232, | |
| "grad_norm": 7.516266345977783, | |
| "learning_rate": 6.476830129583206e-05, | |
| "loss": 0.5713, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.2717842323651452, | |
| "grad_norm": 7.401801109313965, | |
| "learning_rate": 6.434016163555452e-05, | |
| "loss": 0.6164, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.2800829875518671, | |
| "grad_norm": 6.253715515136719, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 0.5391, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.288381742738589, | |
| "grad_norm": 6.459432125091553, | |
| "learning_rate": 6.34804703666072e-05, | |
| "loss": 0.3725, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.2966804979253113, | |
| "grad_norm": 8.033556938171387, | |
| "learning_rate": 6.304898762715186e-05, | |
| "loss": 0.578, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.3049792531120332, | |
| "grad_norm": 6.300682067871094, | |
| "learning_rate": 6.261645954321109e-05, | |
| "loss": 0.4584, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.3132780082987552, | |
| "grad_norm": 6.387180328369141, | |
| "learning_rate": 6.21829207642811e-05, | |
| "loss": 0.4923, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.3215767634854771, | |
| "grad_norm": 7.160585880279541, | |
| "learning_rate": 6.174840602082412e-05, | |
| "loss": 0.5705, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.329875518672199, | |
| "grad_norm": 6.173452854156494, | |
| "learning_rate": 6.131295012148612e-05, | |
| "loss": 0.4962, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3381742738589213, | |
| "grad_norm": 4.847928047180176, | |
| "learning_rate": 6.087658795030837e-05, | |
| "loss": 0.3703, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.3464730290456433, | |
| "grad_norm": 7.544972896575928, | |
| "learning_rate": 6.043935446393294e-05, | |
| "loss": 0.433, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.3547717842323652, | |
| "grad_norm": 6.9235358238220215, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 0.6781, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.3630705394190872, | |
| "grad_norm": 7.10660457611084, | |
| "learning_rate": 5.956241371835312e-05, | |
| "loss": 0.5983, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.3713692946058091, | |
| "grad_norm": 4.3860955238342285, | |
| "learning_rate": 5.912277671020564e-05, | |
| "loss": 0.3219, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.379668049792531, | |
| "grad_norm": 6.816618919372559, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.4972, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.387966804979253, | |
| "grad_norm": 5.665396690368652, | |
| "learning_rate": 5.824134551530783e-05, | |
| "loss": 0.4513, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.396265560165975, | |
| "grad_norm": 8.053374290466309, | |
| "learning_rate": 5.7799621939340896e-05, | |
| "loss": 0.3688, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.404564315352697, | |
| "grad_norm": 5.979791164398193, | |
| "learning_rate": 5.735727354158581e-05, | |
| "loss": 0.3598, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.412863070539419, | |
| "grad_norm": 4.926863193511963, | |
| "learning_rate": 5.691433575823666e-05, | |
| "loss": 0.3195, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4211618257261411, | |
| "grad_norm": 6.836885452270508, | |
| "learning_rate": 5.6470844072702764e-05, | |
| "loss": 0.6218, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.429460580912863, | |
| "grad_norm": 7.920914649963379, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 0.4448, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.437759336099585, | |
| "grad_norm": 3.9926419258117676, | |
| "learning_rate": 5.55823411477354e-05, | |
| "loss": 0.208, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.446058091286307, | |
| "grad_norm": 5.798760414123535, | |
| "learning_rate": 5.513740108559622e-05, | |
| "loss": 0.3248, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.454356846473029, | |
| "grad_norm": 7.048202991485596, | |
| "learning_rate": 5.469204947015897e-05, | |
| "loss": 0.4915, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.4626556016597512, | |
| "grad_norm": 4.450428485870361, | |
| "learning_rate": 5.424632197820324e-05, | |
| "loss": 0.2565, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.4709543568464731, | |
| "grad_norm": 7.559440612792969, | |
| "learning_rate": 5.3800254316619806e-05, | |
| "loss": 0.5286, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.479253112033195, | |
| "grad_norm": 7.2518086433410645, | |
| "learning_rate": 5.335388221955012e-05, | |
| "loss": 0.5451, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.487551867219917, | |
| "grad_norm": 5.655428886413574, | |
| "learning_rate": 5.290724144552379e-05, | |
| "loss": 0.3642, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.495850622406639, | |
| "grad_norm": 6.303919792175293, | |
| "learning_rate": 5.246036777459391e-05, | |
| "loss": 0.4304, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.504149377593361, | |
| "grad_norm": 9.382453918457031, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 0.5932, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.512448132780083, | |
| "grad_norm": 5.657342910766602, | |
| "learning_rate": 5.1566064952654014e-05, | |
| "loss": 0.431, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.520746887966805, | |
| "grad_norm": 5.608704566955566, | |
| "learning_rate": 5.1118707443563665e-05, | |
| "loss": 0.4258, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.5290456431535269, | |
| "grad_norm": 6.01896333694458, | |
| "learning_rate": 5.0671260315669875e-05, | |
| "loss": 0.3907, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.5373443983402488, | |
| "grad_norm": 8.329726219177246, | |
| "learning_rate": 5.022375941362217e-05, | |
| "loss": 0.3328, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5456431535269708, | |
| "grad_norm": 6.260307788848877, | |
| "learning_rate": 4.977624058637783e-05, | |
| "loss": 0.3374, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.553941908713693, | |
| "grad_norm": 6.407407760620117, | |
| "learning_rate": 4.9328739684330137e-05, | |
| "loss": 0.432, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.562240663900415, | |
| "grad_norm": 5.673023700714111, | |
| "learning_rate": 4.8881292556436354e-05, | |
| "loss": 0.2581, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.570539419087137, | |
| "grad_norm": 6.290989875793457, | |
| "learning_rate": 4.8433935047346e-05, | |
| "loss": 0.387, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.578838174273859, | |
| "grad_norm": 5.269549369812012, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 0.3035, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.587136929460581, | |
| "grad_norm": 7.64688777923584, | |
| "learning_rate": 4.7539632225406095e-05, | |
| "loss": 0.3908, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.595435684647303, | |
| "grad_norm": 11.037405014038086, | |
| "learning_rate": 4.709275855447621e-05, | |
| "loss": 0.8535, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.603734439834025, | |
| "grad_norm": 5.834881782531738, | |
| "learning_rate": 4.6646117780449876e-05, | |
| "loss": 0.2963, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.612033195020747, | |
| "grad_norm": 5.2544403076171875, | |
| "learning_rate": 4.6199745683380206e-05, | |
| "loss": 0.2959, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.620331950207469, | |
| "grad_norm": 8.141179084777832, | |
| "learning_rate": 4.5753678021796755e-05, | |
| "loss": 0.634, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.6286307053941909, | |
| "grad_norm": 4.7402191162109375, | |
| "learning_rate": 4.530795052984104e-05, | |
| "loss": 0.2887, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.6369294605809128, | |
| "grad_norm": 6.756740093231201, | |
| "learning_rate": 4.48625989144038e-05, | |
| "loss": 0.4239, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.6452282157676348, | |
| "grad_norm": 6.466264247894287, | |
| "learning_rate": 4.4417658852264614e-05, | |
| "loss": 0.2381, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.6535269709543567, | |
| "grad_norm": 5.09955358505249, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 0.3166, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.6618257261410787, | |
| "grad_norm": 5.525445461273193, | |
| "learning_rate": 4.352915592729723e-05, | |
| "loss": 0.3016, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6618257261410787, | |
| "eval_loss": 0.6099563241004944, | |
| "eval_runtime": 7.1625, | |
| "eval_samples_per_second": 28.342, | |
| "eval_steps_per_second": 7.12, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6701244813278007, | |
| "grad_norm": 7.207304954528809, | |
| "learning_rate": 4.308566424176336e-05, | |
| "loss": 0.5757, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.6784232365145229, | |
| "grad_norm": 8.592957496643066, | |
| "learning_rate": 4.264272645841419e-05, | |
| "loss": 0.487, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.6867219917012448, | |
| "grad_norm": 4.523548603057861, | |
| "learning_rate": 4.2200378060659116e-05, | |
| "loss": 0.1985, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.6950207468879668, | |
| "grad_norm": 3.577657699584961, | |
| "learning_rate": 4.1758654484692186e-05, | |
| "loss": 0.217, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.703319502074689, | |
| "grad_norm": 5.221662998199463, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.3242, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.711618257261411, | |
| "grad_norm": 4.403975009918213, | |
| "learning_rate": 4.087722328979438e-05, | |
| "loss": 0.2363, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.7199170124481329, | |
| "grad_norm": 4.955993175506592, | |
| "learning_rate": 4.043758628164688e-05, | |
| "loss": 0.4589, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.7282157676348548, | |
| "grad_norm": 5.290609836578369, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 0.304, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.7365145228215768, | |
| "grad_norm": 4.659958362579346, | |
| "learning_rate": 3.956064553606708e-05, | |
| "loss": 0.1433, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.7448132780082988, | |
| "grad_norm": 6.113132953643799, | |
| "learning_rate": 3.912341204969164e-05, | |
| "loss": 0.2607, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.7531120331950207, | |
| "grad_norm": 7.161942005157471, | |
| "learning_rate": 3.86870498785139e-05, | |
| "loss": 0.5274, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.7614107883817427, | |
| "grad_norm": 7.1513671875, | |
| "learning_rate": 3.825159397917589e-05, | |
| "loss": 0.4634, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.7697095435684647, | |
| "grad_norm": 5.32570219039917, | |
| "learning_rate": 3.781707923571891e-05, | |
| "loss": 0.1875, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.7780082987551866, | |
| "grad_norm": 4.187331199645996, | |
| "learning_rate": 3.738354045678891e-05, | |
| "loss": 0.1516, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.7863070539419086, | |
| "grad_norm": 7.296047687530518, | |
| "learning_rate": 3.695101237284815e-05, | |
| "loss": 0.3847, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.7946058091286305, | |
| "grad_norm": 5.951739311218262, | |
| "learning_rate": 3.651952963339282e-05, | |
| "loss": 0.2656, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.8029045643153527, | |
| "grad_norm": 6.873353958129883, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 0.165, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.8112033195020747, | |
| "grad_norm": 6.497419834136963, | |
| "learning_rate": 3.5659838364445505e-05, | |
| "loss": 0.2504, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.8195020746887967, | |
| "grad_norm": 7.605969429016113, | |
| "learning_rate": 3.523169870416795e-05, | |
| "loss": 0.3794, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.8278008298755186, | |
| "grad_norm": 7.117875576019287, | |
| "learning_rate": 3.480474212128766e-05, | |
| "loss": 0.3687, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8360995850622408, | |
| "grad_norm": 4.442015647888184, | |
| "learning_rate": 3.4379002818972124e-05, | |
| "loss": 0.2042, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.8443983402489628, | |
| "grad_norm": 8.238639831542969, | |
| "learning_rate": 3.3954514902873425e-05, | |
| "loss": 0.5648, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.8526970954356847, | |
| "grad_norm": 7.620614051818848, | |
| "learning_rate": 3.3531312378396026e-05, | |
| "loss": 0.3973, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.8609958506224067, | |
| "grad_norm": 5.214080333709717, | |
| "learning_rate": 3.310942914797265e-05, | |
| "loss": 0.2133, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.8692946058091287, | |
| "grad_norm": 5.911538600921631, | |
| "learning_rate": 3.2688899008348386e-05, | |
| "loss": 0.219, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.8775933609958506, | |
| "grad_norm": 10.658268928527832, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 0.338, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.8858921161825726, | |
| "grad_norm": 6.180187702178955, | |
| "learning_rate": 3.185203264380338e-05, | |
| "loss": 0.2234, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.8941908713692945, | |
| "grad_norm": 5.702350616455078, | |
| "learning_rate": 3.143576345961132e-05, | |
| "loss": 0.2215, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.9024896265560165, | |
| "grad_norm": 2.7787413597106934, | |
| "learning_rate": 3.1020981442305184e-05, | |
| "loss": 0.0856, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.9107883817427385, | |
| "grad_norm": 5.721746921539307, | |
| "learning_rate": 3.060771981975726e-05, | |
| "loss": 0.3339, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.9190871369294604, | |
| "grad_norm": 6.0323567390441895, | |
| "learning_rate": 3.019601169804216e-05, | |
| "loss": 0.1757, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.9273858921161826, | |
| "grad_norm": 8.973467826843262, | |
| "learning_rate": 2.978589005878476e-05, | |
| "loss": 0.262, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.9356846473029046, | |
| "grad_norm": 3.4507062435150146, | |
| "learning_rate": 2.9377387756517982e-05, | |
| "loss": 0.1235, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.9439834024896265, | |
| "grad_norm": 5.919799327850342, | |
| "learning_rate": 2.897053751605093e-05, | |
| "loss": 0.2862, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.9522821576763485, | |
| "grad_norm": 4.9692487716674805, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 0.1764, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.9605809128630707, | |
| "grad_norm": 5.740837574005127, | |
| "learning_rate": 2.8161923455414367e-05, | |
| "loss": 0.1705, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.9688796680497926, | |
| "grad_norm": 5.595933437347412, | |
| "learning_rate": 2.776022441270295e-05, | |
| "loss": 0.2613, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.9771784232365146, | |
| "grad_norm": 4.591209888458252, | |
| "learning_rate": 2.7360306981518146e-05, | |
| "loss": 0.1906, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.9854771784232366, | |
| "grad_norm": 7.699284553527832, | |
| "learning_rate": 2.6962203198941587e-05, | |
| "loss": 0.4312, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.9937759336099585, | |
| "grad_norm": 6.995236873626709, | |
| "learning_rate": 2.656594495676482e-05, | |
| "loss": 0.3351, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.004149377593361, | |
| "grad_norm": 6.956718921661377, | |
| "learning_rate": 2.6171563998934605e-05, | |
| "loss": 0.2213, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.012448132780083, | |
| "grad_norm": 3.839200258255005, | |
| "learning_rate": 2.5779091919009877e-05, | |
| "loss": 0.2139, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.020746887966805, | |
| "grad_norm": 1.65491783618927, | |
| "learning_rate": 2.5388560157630765e-05, | |
| "loss": 0.0616, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.029045643153527, | |
| "grad_norm": 3.041247606277466, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.0922, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.037344398340249, | |
| "grad_norm": 1.837149739265442, | |
| "learning_rate": 2.461344257337662e-05, | |
| "loss": 0.0552, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.045643153526971, | |
| "grad_norm": 3.2361204624176025, | |
| "learning_rate": 2.422891884458241e-05, | |
| "loss": 0.0536, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.0539419087136928, | |
| "grad_norm": 4.026633262634277, | |
| "learning_rate": 2.3846459617521128e-05, | |
| "loss": 0.2692, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.0622406639004147, | |
| "grad_norm": 3.3415215015411377, | |
| "learning_rate": 2.346609553071093e-05, | |
| "loss": 0.0902, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.070539419087137, | |
| "grad_norm": 0.9877287745475769, | |
| "learning_rate": 2.308785705482982e-05, | |
| "loss": 0.0222, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.078838174273859, | |
| "grad_norm": 1.6253547668457031, | |
| "learning_rate": 2.2711774490274766e-05, | |
| "loss": 0.0311, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.078838174273859, | |
| "eval_loss": 0.4097523093223572, | |
| "eval_runtime": 7.1697, | |
| "eval_samples_per_second": 28.314, | |
| "eval_steps_per_second": 7.113, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.087136929460581, | |
| "grad_norm": 4.143763065338135, | |
| "learning_rate": 2.233787796473432e-05, | |
| "loss": 0.0556, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.095435684647303, | |
| "grad_norm": 4.3261847496032715, | |
| "learning_rate": 2.1966197430775053e-05, | |
| "loss": 0.1249, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.103734439834025, | |
| "grad_norm": 4.310375690460205, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 0.0994, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.112033195020747, | |
| "grad_norm": 1.23174250125885, | |
| "learning_rate": 2.122960325787432e-05, | |
| "loss": 0.0315, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.120331950207469, | |
| "grad_norm": 1.8426965475082397, | |
| "learning_rate": 2.086474862693244e-05, | |
| "loss": 0.0491, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.128630705394191, | |
| "grad_norm": 3.062448501586914, | |
| "learning_rate": 2.050222799884387e-05, | |
| "loss": 0.0968, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.136929460580913, | |
| "grad_norm": 2.411597967147827, | |
| "learning_rate": 2.0142070414860704e-05, | |
| "loss": 0.0283, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.145228215767635, | |
| "grad_norm": 2.3557469844818115, | |
| "learning_rate": 1.9784304726933383e-05, | |
| "loss": 0.0897, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.1535269709543567, | |
| "grad_norm": 8.79699420928955, | |
| "learning_rate": 1.942895959539939e-05, | |
| "loss": 0.0328, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.1618257261410787, | |
| "grad_norm": 1.4276204109191895, | |
| "learning_rate": 1.9076063486687256e-05, | |
| "loss": 0.016, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.1701244813278007, | |
| "grad_norm": 2.912771701812744, | |
| "learning_rate": 1.8725644671036126e-05, | |
| "loss": 0.1233, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.1784232365145226, | |
| "grad_norm": 4.887458801269531, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 0.0717, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.186721991701245, | |
| "grad_norm": 1.242602825164795, | |
| "learning_rate": 1.803235100535452e-05, | |
| "loss": 0.0208, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.195020746887967, | |
| "grad_norm": 0.9755971431732178, | |
| "learning_rate": 1.7689531694552863e-05, | |
| "loss": 0.0192, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.203319502074689, | |
| "grad_norm": 2.5533509254455566, | |
| "learning_rate": 1.734930075082076e-05, | |
| "loss": 0.0384, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.211618257261411, | |
| "grad_norm": 0.9231969118118286, | |
| "learning_rate": 1.7011685429800595e-05, | |
| "loss": 0.0152, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.219917012448133, | |
| "grad_norm": 1.7765648365020752, | |
| "learning_rate": 1.6676712777599273e-05, | |
| "loss": 0.041, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.228215767634855, | |
| "grad_norm": 2.0227739810943604, | |
| "learning_rate": 1.6344409628621484e-05, | |
| "loss": 0.1056, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.236514522821577, | |
| "grad_norm": 1.8746280670166016, | |
| "learning_rate": 1.6014802603420044e-05, | |
| "loss": 0.0542, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.2448132780082988, | |
| "grad_norm": 1.8291066884994507, | |
| "learning_rate": 1.5687918106563326e-05, | |
| "loss": 0.0206, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.2531120331950207, | |
| "grad_norm": 7.360498905181885, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 0.1436, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.2614107883817427, | |
| "grad_norm": 4.359367847442627, | |
| "learning_rate": 1.5042421223561432e-05, | |
| "loss": 0.075, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.2697095435684647, | |
| "grad_norm": 1.2709429264068604, | |
| "learning_rate": 1.4723860547681162e-05, | |
| "loss": 0.0172, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.2780082987551866, | |
| "grad_norm": 3.57598876953125, | |
| "learning_rate": 1.440812581653298e-05, | |
| "loss": 0.0619, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.2863070539419086, | |
| "grad_norm": 1.036085605621338, | |
| "learning_rate": 1.4095242323386303e-05, | |
| "loss": 0.0604, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.2946058091286305, | |
| "grad_norm": 2.2307868003845215, | |
| "learning_rate": 1.3785235133100088e-05, | |
| "loss": 0.037, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.3029045643153525, | |
| "grad_norm": 1.6419095993041992, | |
| "learning_rate": 1.3478129080114848e-05, | |
| "loss": 0.0735, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.3112033195020745, | |
| "grad_norm": 2.744889259338379, | |
| "learning_rate": 1.3173948766463145e-05, | |
| "loss": 0.0294, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.3195020746887964, | |
| "grad_norm": 3.0023012161254883, | |
| "learning_rate": 1.2872718559798853e-05, | |
| "loss": 0.0446, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.327800829875519, | |
| "grad_norm": 4.369894027709961, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 0.0467, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.336099585062241, | |
| "grad_norm": 1.3997019529342651, | |
| "learning_rate": 1.2279204754460493e-05, | |
| "loss": 0.0205, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.3443983402489628, | |
| "grad_norm": 3.3227059841156006, | |
| "learning_rate": 1.1986968701726491e-05, | |
| "loss": 0.083, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.3526970954356847, | |
| "grad_norm": 2.5781474113464355, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 0.0682, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.3609958506224067, | |
| "grad_norm": 0.8888638019561768, | |
| "learning_rate": 1.1411655348294247e-05, | |
| "loss": 0.0169, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.3692946058091287, | |
| "grad_norm": 3.076343059539795, | |
| "learning_rate": 1.1128624135511712e-05, | |
| "loss": 0.0746, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.3775933609958506, | |
| "grad_norm": 2.557579755783081, | |
| "learning_rate": 1.0848706879118892e-05, | |
| "loss": 0.0715, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.3858921161825726, | |
| "grad_norm": 3.306203842163086, | |
| "learning_rate": 1.0571926003074561e-05, | |
| "loss": 0.063, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.3941908713692945, | |
| "grad_norm": 1.7703243494033813, | |
| "learning_rate": 1.0298303680084448e-05, | |
| "loss": 0.0303, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.4024896265560165, | |
| "grad_norm": 3.1793651580810547, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 0.0253, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.4107883817427385, | |
| "grad_norm": 2.1019341945648193, | |
| "learning_rate": 9.760622117187235e-06, | |
| "loss": 0.0759, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.4190871369294604, | |
| "grad_norm": 1.9458683729171753, | |
| "learning_rate": 9.496605950541676e-06, | |
| "loss": 0.0259, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.4273858921161824, | |
| "grad_norm": 3.2663235664367676, | |
| "learning_rate": 9.235834480022787e-06, | |
| "loss": 0.071, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.435684647302905, | |
| "grad_norm": 2.9131743907928467, | |
| "learning_rate": 8.978328595834984e-06, | |
| "loss": 0.0265, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.4439834024896268, | |
| "grad_norm": 4.122836589813232, | |
| "learning_rate": 8.724108926578999e-06, | |
| "loss": 0.0449, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.4522821576763487, | |
| "grad_norm": 2.857977867126465, | |
| "learning_rate": 8.473195837599418e-06, | |
| "loss": 0.0248, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.4605809128630707, | |
| "grad_norm": 4.170780181884766, | |
| "learning_rate": 8.225609429353187e-06, | |
| "loss": 0.0373, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.4688796680497926, | |
| "grad_norm": 2.490182638168335, | |
| "learning_rate": 7.981369535799354e-06, | |
| "loss": 0.0751, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.4771784232365146, | |
| "grad_norm": 3.713991165161133, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 0.0964, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.4854771784232366, | |
| "grad_norm": 3.0123841762542725, | |
| "learning_rate": 7.503007286604069e-06, | |
| "loss": 0.0414, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.4937759336099585, | |
| "grad_norm": 4.080419063568115, | |
| "learning_rate": 7.268923252198989e-06, | |
| "loss": 0.1049, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.4937759336099585, | |
| "eval_loss": 0.3370027542114258, | |
| "eval_runtime": 7.1786, | |
| "eval_samples_per_second": 28.278, | |
| "eval_steps_per_second": 7.104, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.5020746887966805, | |
| "grad_norm": 5.102431774139404, | |
| "learning_rate": 7.038262371889159e-06, | |
| "loss": 0.1422, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.5103734439834025, | |
| "grad_norm": 1.8823308944702148, | |
| "learning_rate": 6.811043123742494e-06, | |
| "loss": 0.0404, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.5186721991701244, | |
| "grad_norm": 1.6759247779846191, | |
| "learning_rate": 6.587283710120324e-06, | |
| "loss": 0.0273, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.5269709543568464, | |
| "grad_norm": 1.534498929977417, | |
| "learning_rate": 6.367002056219284e-06, | |
| "loss": 0.0274, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.5352697095435683, | |
| "grad_norm": 1.8221664428710938, | |
| "learning_rate": 6.150215808635335e-06, | |
| "loss": 0.1055, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.5435684647302903, | |
| "grad_norm": 2.2624080181121826, | |
| "learning_rate": 5.936942333950063e-06, | |
| "loss": 0.0522, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.5518672199170123, | |
| "grad_norm": 1.1444363594055176, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 0.0714, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.5601659751037342, | |
| "grad_norm": 1.0299466848373413, | |
| "learning_rate": 5.521001761205441e-06, | |
| "loss": 0.017, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.568464730290456, | |
| "grad_norm": 1.962306022644043, | |
| "learning_rate": 5.318367983829392e-06, | |
| "loss": 0.0308, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.576763485477178, | |
| "grad_norm": 1.333689570426941, | |
| "learning_rate": 5.1193136180493095e-06, | |
| "loss": 0.0488, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.5850622406639006, | |
| "grad_norm": 1.2325903177261353, | |
| "learning_rate": 4.9238546099592e-06, | |
| "loss": 0.0345, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.5933609958506225, | |
| "grad_norm": 1.9375916719436646, | |
| "learning_rate": 4.732006617631729e-06, | |
| "loss": 0.0257, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.6016597510373445, | |
| "grad_norm": 7.875622749328613, | |
| "learning_rate": 4.54378500986381e-06, | |
| "loss": 0.2938, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.6099585062240664, | |
| "grad_norm": 1.0272877216339111, | |
| "learning_rate": 4.3592048649454594e-06, | |
| "loss": 0.0183, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.6182572614107884, | |
| "grad_norm": 3.053288459777832, | |
| "learning_rate": 4.178280969451853e-06, | |
| "loss": 0.0555, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.6265560165975104, | |
| "grad_norm": 1.3870058059692383, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 0.0233, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.6348547717842323, | |
| "grad_norm": 2.4694643020629883, | |
| "learning_rate": 3.827459607381678e-06, | |
| "loss": 0.0883, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.6431535269709543, | |
| "grad_norm": 1.914157748222351, | |
| "learning_rate": 3.6575902448379105e-06, | |
| "loss": 0.0428, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.6514522821576763, | |
| "grad_norm": 2.42271089553833, | |
| "learning_rate": 3.4914333375330898e-06, | |
| "loss": 0.0574, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.659751037344398, | |
| "grad_norm": 1.2127854824066162, | |
| "learning_rate": 3.329002196170816e-06, | |
| "loss": 0.0504, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.66804979253112, | |
| "grad_norm": 2.0530266761779785, | |
| "learning_rate": 3.1703098329864233e-06, | |
| "loss": 0.0504, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.6763485477178426, | |
| "grad_norm": 1.7026387453079224, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 0.1294, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.6846473029045645, | |
| "grad_norm": 2.17067813873291, | |
| "learning_rate": 2.864191991520848e-06, | |
| "loss": 0.0188, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.6929460580912865, | |
| "grad_norm": 0.5803589820861816, | |
| "learning_rate": 2.71679103610738e-06, | |
| "loss": 0.0468, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.7012448132780085, | |
| "grad_norm": 2.2889926433563232, | |
| "learning_rate": 2.573177902642726e-06, | |
| "loss": 0.1123, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.7095435684647304, | |
| "grad_norm": 1.7318047285079956, | |
| "learning_rate": 2.4333640958659143e-06, | |
| "loss": 0.027, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.7178423236514524, | |
| "grad_norm": 2.7389817237854004, | |
| "learning_rate": 2.2973608161547753e-06, | |
| "loss": 0.08, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.7261410788381744, | |
| "grad_norm": 1.5119000673294067, | |
| "learning_rate": 2.1651789586287442e-06, | |
| "loss": 0.1249, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.7344398340248963, | |
| "grad_norm": 0.7453382015228271, | |
| "learning_rate": 2.03682911227599e-06, | |
| "loss": 0.0185, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.7427385892116183, | |
| "grad_norm": 1.6104263067245483, | |
| "learning_rate": 1.9123215591052013e-06, | |
| "loss": 0.0319, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.7510373443983402, | |
| "grad_norm": 2.6384763717651367, | |
| "learning_rate": 1.7916662733218847e-06, | |
| "loss": 0.1004, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.759336099585062, | |
| "grad_norm": 4.27977180480957, | |
| "learning_rate": 1.6748729205293023e-06, | |
| "loss": 0.1176, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 2.767634854771784, | |
| "grad_norm": 2.4761962890625, | |
| "learning_rate": 1.5619508569542363e-06, | |
| "loss": 0.0396, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 2.775933609958506, | |
| "grad_norm": 2.1308181285858154, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 0.0385, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 2.784232365145228, | |
| "grad_norm": 1.79630446434021, | |
| "learning_rate": 1.3477564710088098e-06, | |
| "loss": 0.0193, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.79253112033195, | |
| "grad_norm": 6.298658847808838, | |
| "learning_rate": 1.2465013075879883e-06, | |
| "loss": 0.0399, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 2.800829875518672, | |
| "grad_norm": 1.5001986026763916, | |
| "learning_rate": 1.1491517499091498e-06, | |
| "loss": 0.0225, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 2.809128630705394, | |
| "grad_norm": 1.3318809270858765, | |
| "learning_rate": 1.055715596571405e-06, | |
| "loss": 0.0212, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 2.817427385892116, | |
| "grad_norm": 0.8632059693336487, | |
| "learning_rate": 9.662003326740166e-07, | |
| "loss": 0.0136, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 2.825726141078838, | |
| "grad_norm": 2.7577383518218994, | |
| "learning_rate": 8.806131292167618e-07, | |
| "loss": 0.068, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.8340248962655603, | |
| "grad_norm": 1.1640796661376953, | |
| "learning_rate": 7.989608425254924e-07, | |
| "loss": 0.0268, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 2.8423236514522823, | |
| "grad_norm": 2.3888018131256104, | |
| "learning_rate": 7.212500137028788e-07, | |
| "loss": 0.0754, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 2.8506224066390042, | |
| "grad_norm": 3.372357130050659, | |
| "learning_rate": 6.474868681043578e-07, | |
| "loss": 0.1583, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 2.858921161825726, | |
| "grad_norm": 0.9823101162910461, | |
| "learning_rate": 5.776773148394976e-07, | |
| "loss": 0.0163, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 2.867219917012448, | |
| "grad_norm": 1.4585671424865723, | |
| "learning_rate": 5.118269462985714e-07, | |
| "loss": 0.0248, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.87551867219917, | |
| "grad_norm": 2.173081636428833, | |
| "learning_rate": 4.4994103770457653e-07, | |
| "loss": 0.0443, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 2.883817427385892, | |
| "grad_norm": 1.831309199333191, | |
| "learning_rate": 3.920245466906391e-07, | |
| "loss": 0.0713, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 2.892116182572614, | |
| "grad_norm": 1.361579179763794, | |
| "learning_rate": 3.380821129028489e-07, | |
| "loss": 0.0208, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 2.900414937759336, | |
| "grad_norm": 1.126133680343628, | |
| "learning_rate": 2.8811805762860576e-07, | |
| "loss": 0.01, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 2.908713692946058, | |
| "grad_norm": 0.6526811122894287, | |
| "learning_rate": 2.421363834504087e-07, | |
| "loss": 0.0093, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.908713692946058, | |
| "eval_loss": 0.32943063974380493, | |
| "eval_runtime": 7.1396, | |
| "eval_samples_per_second": 28.433, | |
| "eval_steps_per_second": 7.143, | |
| "step": 350 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 361, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.390780807643136e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |