{ "best_metric": 0.47966957092285156, "best_model_checkpoint": "nyt_ingredients-tagger-paraphrase-MiniLM-L3-v2/checkpoint-24000", "epoch": 10.0, "eval_steps": 1000, "global_step": 50510, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.1, "learning_rate": 4.9505048505246485e-05, "loss": 1.1204, "step": 500 }, { "epoch": 0.2, "learning_rate": 4.9010097010492975e-05, "loss": 0.7174, "step": 1000 }, { "epoch": 0.2, "eval_COMMENT": { "f1": 0.5647948868453662, "number": 6824, "precision": 0.5088729580444236, "recall": 0.6345252051582649 }, "eval_NAME": { "f1": 0.7765338110165697, "number": 8803, "precision": 0.7654784240150094, "recall": 0.7879132114052028 }, "eval_QTY": { "f1": 0.9714048901782015, "number": 7168, "precision": 0.9619699042407661, "recall": 0.9810267857142857 }, "eval_RANGE_END": { "f1": 0.0, "number": 82, "precision": 0.0, "recall": 0.0 }, "eval_UNIT": { "f1": 0.945750755794424, "number": 5755, "precision": 0.9151633349585568, "recall": 0.9784535186794092 }, "eval_loss": 0.656587541103363, "eval_overall_accuracy": 0.7841206721853672, "eval_overall_f1": 0.8021117908321461, "eval_overall_precision": 0.7710823961589276, "eval_overall_recall": 0.8357432243643476, "eval_runtime": 7.1565, "eval_samples_per_second": 1188.573, "eval_steps_per_second": 37.169, "step": 1000 }, { "epoch": 0.3, "learning_rate": 4.851514551573946e-05, "loss": 0.6502, "step": 1500 }, { "epoch": 0.4, "learning_rate": 4.802019402098595e-05, "loss": 0.6076, "step": 2000 }, { "epoch": 0.4, "eval_COMMENT": { "f1": 0.5854632587859425, "number": 6824, "precision": 0.5363414634146342, "recall": 0.6444900351699883 }, "eval_NAME": { "f1": 0.7806724071738058, "number": 8803, "precision": 0.7547189819724284, "recall": 0.8084743837328183 }, "eval_QTY": { "f1": 0.9778333680772705, "number": 7168, "precision": 0.974110480409802, "recall": 0.9815848214285714 }, "eval_RANGE_END": { "f1": 0.23529411764705882, "number": 82, "precision": 0.6, "recall": 0.14634146341463414 }, "eval_UNIT": { "f1": 0.9454667112746737, "number": 5755, "precision": 0.9114658925979681, "recall": 0.9821025195482189 }, "eval_loss": 0.584534227848053, "eval_overall_accuracy": 0.7988032194543752, "eval_overall_f1": 0.811141258834958, "eval_overall_precision": 0.7792688421188132, "eval_overall_recall": 0.8457320480581167, "eval_runtime": 6.3972, "eval_samples_per_second": 1329.645, "eval_steps_per_second": 41.581, "step": 2000 }, { "epoch": 0.49, "learning_rate": 4.752524252623243e-05, "loss": 0.5843, "step": 2500 }, { "epoch": 0.59, "learning_rate": 4.703029103147892e-05, "loss": 0.5777, "step": 3000 }, { "epoch": 0.59, "eval_COMMENT": { "f1": 0.6086500655307995, "number": 6824, "precision": 0.55049786628734, "recall": 0.6805392731535757 }, "eval_NAME": { "f1": 0.7919685823330935, "number": 8803, "precision": 0.7717766278568349, "recall": 0.8132454844939225 }, "eval_QTY": { "f1": 0.9790267875683533, "number": 7168, "precision": 0.9715620277510647, "recall": 0.9866071428571429 }, "eval_RANGE_END": { "f1": 0.5133689839572192, "number": 82, "precision": 0.45714285714285713, "recall": 0.5853658536585366 }, "eval_UNIT": { "f1": 0.9490924440692275, "number": 5755, "precision": 0.9229885057471264, "recall": 0.9767158992180712 }, "eval_loss": 0.5503401756286621, "eval_overall_accuracy": 0.805647427686543, "eval_overall_f1": 0.820622555083754, "eval_overall_precision": 0.7870198165843648, "eval_overall_recall": 0.8572226879016485, "eval_runtime": 8.9283, "eval_samples_per_second": 952.704, "eval_steps_per_second": 29.793, "step": 3000 }, { "epoch": 0.69, "learning_rate": 4.65353395367254e-05, "loss": 0.5581, "step": 3500 }, { "epoch": 0.79, "learning_rate": 4.6040388041971886e-05, "loss": 0.553, "step": 4000 }, { "epoch": 0.79, "eval_COMMENT": { "f1": 0.6282245827010623, "number": 6824, "precision": 0.5713428537141486, "recall": 0.6976846424384525 }, "eval_NAME": { "f1": 0.7973033207042568, "number": 8803, "precision": 0.7823947512301804, "recall": 0.812791093945246 }, "eval_QTY": { "f1": 0.9824755835699939, "number": 7168, "precision": 0.9756500206355757, "recall": 0.9893973214285714 }, "eval_RANGE_END": { "f1": 0.5595238095238095, "number": 82, "precision": 0.5465116279069767, "recall": 0.573170731707317 }, "eval_UNIT": { "f1": 0.9491468718634996, "number": 5755, "precision": 0.9150137074665377, "recall": 0.9859252823631625 }, "eval_loss": 0.5342816710472107, "eval_overall_accuracy": 0.8121666316171832, "eval_overall_f1": 0.8289142895451346, "eval_overall_precision": 0.7968357285557776, "eval_overall_recall": 0.8636839899413243, "eval_runtime": 7.0508, "eval_samples_per_second": 1206.385, "eval_steps_per_second": 37.726, "step": 4000 }, { "epoch": 0.89, "learning_rate": 4.5545436547218376e-05, "loss": 0.5436, "step": 4500 }, { "epoch": 0.99, "learning_rate": 4.505048505246486e-05, "loss": 0.5407, "step": 5000 }, { "epoch": 0.99, "eval_COMMENT": { "f1": 0.6475744396119103, "number": 6824, "precision": 0.5958625785001848, "recall": 0.7091148886283705 }, "eval_NAME": { "f1": 0.8001564333202973, "number": 8803, "precision": 0.787269129287599, "recall": 0.8134726797682609 }, "eval_QTY": { "f1": 0.9841955023323818, "number": 7168, "precision": 0.9823488533703961, "recall": 0.9860491071428571 }, "eval_RANGE_END": { "f1": 0.626865671641791, "number": 82, "precision": 0.5294117647058824, "recall": 0.7682926829268293 }, "eval_UNIT": { "f1": 0.9520282780676653, "number": 5755, "precision": 0.9231271421576628, "recall": 0.9827975673327541 }, "eval_loss": 0.518293023109436, "eval_overall_accuracy": 0.8185520102471945, "eval_overall_f1": 0.8361275088547816, "eval_overall_precision": 0.8085002283253963, "eval_overall_recall": 0.8657096954456552, "eval_runtime": 7.4009, "eval_samples_per_second": 1149.315, "eval_steps_per_second": 35.941, "step": 5000 }, { "epoch": 1.09, "learning_rate": 4.455553355771135e-05, "loss": 0.5231, "step": 5500 }, { "epoch": 1.19, "learning_rate": 4.406058206295783e-05, "loss": 0.5109, "step": 6000 }, { "epoch": 1.19, "eval_COMMENT": { "f1": 0.6651526284149355, "number": 6824, "precision": 0.611145206824598, "recall": 0.729630715123095 }, "eval_NAME": { "f1": 0.803545545768759, "number": 8803, "precision": 0.788943623426382, "recall": 0.8186981710780415 }, "eval_QTY": { "f1": 0.9848337275636566, "number": 7168, "precision": 0.9822370247016375, "recall": 0.9874441964285714 }, "eval_RANGE_END": { "f1": 0.653061224489796, "number": 82, "precision": 0.5614035087719298, "recall": 0.7804878048780488 }, "eval_UNIT": { "f1": 0.9527340129749768, "number": 5755, "precision": 0.924762839385018, "recall": 0.9824500434404866 }, "eval_loss": 0.5209813714027405, "eval_overall_accuracy": 0.8200623243542929, "eval_overall_f1": 0.841881782031408, "eval_overall_precision": 0.8133220471415549, "eval_overall_recall": 0.8725202570550433, "eval_runtime": 8.4151, "eval_samples_per_second": 1010.797, "eval_steps_per_second": 31.61, "step": 6000 }, { "epoch": 1.29, "learning_rate": 4.356563056820432e-05, "loss": 0.5232, "step": 6500 }, { "epoch": 1.39, "learning_rate": 4.3070679073450804e-05, "loss": 0.5125, "step": 7000 }, { "epoch": 1.39, "eval_COMMENT": { "f1": 0.6596956900596554, "number": 6824, "precision": 0.607906114885732, "recall": 0.7211313012895663 }, "eval_NAME": { "f1": 0.8034675615212529, "number": 8803, "precision": 0.7913407513495648, "recall": 0.8159718277859821 }, "eval_QTY": { "f1": 0.9845339278249966, "number": 7168, "precision": 0.983300862788756, "recall": 0.9857700892857143 }, "eval_RANGE_END": { "f1": 0.641711229946524, "number": 82, "precision": 0.5714285714285714, "recall": 0.7317073170731707 }, "eval_UNIT": { "f1": 0.952638700947226, "number": 5755, "precision": 0.9279947273026857, "recall": 0.978627280625543 }, "eval_loss": 0.5072416663169861, "eval_overall_accuracy": 0.8215152847611218, "eval_overall_f1": 0.8404435129470623, "eval_overall_precision": 0.8142932005764444, "eval_overall_recall": 0.8683291422184968, "eval_runtime": 6.3336, "eval_samples_per_second": 1342.988, "eval_steps_per_second": 41.998, "step": 7000 }, { "epoch": 1.48, "learning_rate": 4.257572757869729e-05, "loss": 0.5084, "step": 7500 }, { "epoch": 1.58, "learning_rate": 4.208077608394378e-05, "loss": 0.4967, "step": 8000 }, { "epoch": 1.58, "eval_COMMENT": { "f1": 0.6653271716562856, "number": 6824, "precision": 0.6126803996546195, "recall": 0.7278722157092614 }, "eval_NAME": { "f1": 0.8041098197742967, "number": 8803, "precision": 0.7949600355239786, "recall": 0.8134726797682609 }, "eval_QTY": { "f1": 0.9852011394427846, "number": 7168, "precision": 0.9813148788927336, "recall": 0.9891183035714286 }, "eval_RANGE_END": { "f1": 0.6431718061674009, "number": 82, "precision": 0.503448275862069, "recall": 0.8902439024390244 }, "eval_UNIT": { "f1": 0.9540510918135063, "number": 5755, "precision": 0.9266295447101212, "recall": 0.9831450912250217 }, "eval_loss": 0.5088181495666504, "eval_overall_accuracy": 0.8218402890626494, "eval_overall_f1": 0.8425442817824156, "eval_overall_precision": 0.815566669935602, "eval_overall_recall": 0.871367700474993, "eval_runtime": 8.617, "eval_samples_per_second": 987.123, "eval_steps_per_second": 30.869, "step": 8000 }, { "epoch": 1.68, "learning_rate": 4.158582458919026e-05, "loss": 0.5024, "step": 8500 }, { "epoch": 1.78, "learning_rate": 4.109087309443675e-05, "loss": 0.5121, "step": 9000 }, { "epoch": 1.78, "eval_COMMENT": { "f1": 0.6751009421265142, "number": 6824, "precision": 0.6241911398705824, "recall": 0.735052754982415 }, "eval_NAME": { "f1": 0.803609223571349, "number": 8803, "precision": 0.7883291443558081, "recall": 0.8194933545382256 }, "eval_QTY": { "f1": 0.9847892827239743, "number": 7168, "precision": 0.9850642099385818, "recall": 0.9845145089285714 }, "eval_RANGE_END": { "f1": 0.6346153846153846, "number": 82, "precision": 0.5238095238095238, "recall": 0.8048780487804879 }, "eval_UNIT": { "f1": 0.9542967763656258, "number": 5755, "precision": 0.9253999347045381, "recall": 0.9850564726324935 }, "eval_loss": 0.502194344997406, "eval_overall_accuracy": 0.824765327776397, "eval_overall_f1": 0.8448383557018655, "eval_overall_precision": 0.8176322582753325, "eval_overall_recall": 0.8739172953338922, "eval_runtime": 8.4909, "eval_samples_per_second": 1001.783, "eval_steps_per_second": 31.328, "step": 9000 }, { "epoch": 1.88, "learning_rate": 4.059592159968323e-05, "loss": 0.4946, "step": 9500 }, { "epoch": 1.98, "learning_rate": 4.010097010492972e-05, "loss": 0.5008, "step": 10000 }, { "epoch": 1.98, "eval_COMMENT": { "f1": 0.6771771771771772, "number": 6824, "precision": 0.6337506387327542, "recall": 0.7269929660023446 }, "eval_NAME": { "f1": 0.8049774008146867, "number": 8803, "precision": 0.7910726036411494, "recall": 0.8193797569010565 }, "eval_QTY": { "f1": 0.9824440619621343, "number": 7168, "precision": 0.9698246567894522, "recall": 0.9953962053571429 }, "eval_RANGE_END": { "f1": 0.625, "number": 82, "precision": 0.47468354430379744, "recall": 0.9146341463414634 }, "eval_UNIT": { "f1": 0.9544190749010025, "number": 5755, "precision": 0.9263984298331698, "recall": 0.9841876629018245 }, "eval_loss": 0.4997539520263672, "eval_overall_accuracy": 0.8250138604775652, "eval_overall_f1": 0.8461161686962692, "eval_overall_precision": 0.8192313982011448, "eval_overall_recall": 0.874825370215144, "eval_runtime": 6.4794, "eval_samples_per_second": 1312.774, "eval_steps_per_second": 41.053, "step": 10000 }, { "epoch": 2.08, "learning_rate": 3.9606018610176205e-05, "loss": 0.4788, "step": 10500 }, { "epoch": 2.18, "learning_rate": 3.911106711542269e-05, "loss": 0.4863, "step": 11000 }, { "epoch": 2.18, "eval_COMMENT": { "f1": 0.6816733337853415, "number": 6824, "precision": 0.6343217665615142, "recall": 0.736664712778429 }, "eval_NAME": { "f1": 0.8056887897378694, "number": 8803, "precision": 0.7913881888901063, "recall": 0.820515733272748 }, "eval_QTY": { "f1": 0.9852848873701096, "number": 7168, "precision": 0.9850787895690978, "recall": 0.9854910714285714 }, "eval_RANGE_END": { "f1": 0.6666666666666666, "number": 82, "precision": 0.6122448979591837, "recall": 0.7317073170731707 }, "eval_UNIT": { "f1": 0.9536959084020878, "number": 5755, "precision": 0.9250367466927977, "recall": 0.9841876629018245 }, "eval_loss": 0.49540793895721436, "eval_overall_accuracy": 0.8272888905882578, "eval_overall_f1": 0.8476538695917124, "eval_overall_precision": 0.8224280646432794, "eval_overall_recall": 0.8744761106454317, "eval_runtime": 9.0889, "eval_samples_per_second": 935.867, "eval_steps_per_second": 29.266, "step": 11000 }, { "epoch": 2.28, "learning_rate": 3.861611562066917e-05, "loss": 0.4922, "step": 11500 }, { "epoch": 2.38, "learning_rate": 3.812116412591566e-05, "loss": 0.4893, "step": 12000 }, { "epoch": 2.38, "eval_COMMENT": { "f1": 0.6847270771712826, "number": 6824, "precision": 0.6325136612021858, "recall": 0.7463364595545134 }, "eval_NAME": { "f1": 0.8033142412278262, "number": 8803, "precision": 0.7868191721132898, "recall": 0.820515733272748 }, "eval_QTY": { "f1": 0.9853482397055761, "number": 7168, "precision": 0.9809207797594359, "recall": 0.9898158482142857 }, "eval_RANGE_END": { "f1": 0.6491228070175438, "number": 82, "precision": 0.5068493150684932, "recall": 0.9024390243902439 }, "eval_UNIT": { "f1": 0.9545187053383775, "number": 5755, "precision": 0.9245928338762215, "recall": 0.9864465682015638 }, "eval_loss": 0.49466800689697266, "eval_overall_accuracy": 0.8257212227808898, "eval_overall_f1": 0.8474479228061903, "eval_overall_precision": 0.818249813014211, "eval_overall_recall": 0.878806929309863, "eval_runtime": 7.6267, "eval_samples_per_second": 1115.288, "eval_steps_per_second": 34.877, "step": 12000 }, { "epoch": 2.47, "learning_rate": 3.762621263116215e-05, "loss": 0.4767, "step": 12500 }, { "epoch": 2.57, "learning_rate": 3.7131261136408633e-05, "loss": 0.489, "step": 13000 }, { "epoch": 2.57, "eval_COMMENT": { "f1": 0.6822220714915552, "number": 6824, "precision": 0.635054931178179, "recall": 0.736957796014068 }, "eval_NAME": { "f1": 0.8048549635320973, "number": 8803, "precision": 0.7892552959161389, "recall": 0.8210837214585937 }, "eval_QTY": { "f1": 0.9860937282714505, "number": 7168, "precision": 0.9829498197948433, "recall": 0.9892578125 }, "eval_RANGE_END": { "f1": 0.6574074074074074, "number": 82, "precision": 0.5298507462686567, "recall": 0.8658536585365854 }, "eval_UNIT": { "f1": 0.9543087518898035, "number": 5755, "precision": 0.9235896602178507, "recall": 0.9871416159860991 }, "eval_loss": 0.48692840337753296, "eval_overall_accuracy": 0.8271550652876288, "eval_overall_f1": 0.8478583975138494, "eval_overall_precision": 0.8209052851909995, "eval_overall_recall": 0.8766415199776474, "eval_runtime": 6.7601, "eval_samples_per_second": 1258.259, "eval_steps_per_second": 39.348, "step": 13000 }, { "epoch": 2.67, "learning_rate": 3.663630964165512e-05, "loss": 0.4794, "step": 13500 }, { "epoch": 2.77, "learning_rate": 3.6141358146901606e-05, "loss": 0.4691, "step": 14000 }, { "epoch": 2.77, "eval_COMMENT": { "f1": 0.6935899173329719, "number": 6824, "precision": 0.6450718427022939, "recall": 0.75 }, "eval_NAME": { "f1": 0.8057417336907953, "number": 8803, "precision": 0.7925502692011867, "recall": 0.8193797569010565 }, "eval_QTY": { "f1": 0.9854750156369451, "number": 7168, "precision": 0.9818584683561834, "recall": 0.9891183035714286 }, "eval_RANGE_END": { "f1": 0.6820276497695853, "number": 82, "precision": 0.5481481481481482, "recall": 0.9024390243902439 }, "eval_UNIT": { "f1": 0.9554861052453754, "number": 5755, "precision": 0.9296515450361604, "recall": 0.9827975673327541 }, "eval_loss": 0.4904623329639435, "eval_overall_accuracy": 0.8282639034928403, "eval_overall_f1": 0.8510328725870033, "eval_overall_precision": 0.8252994257588186, "eval_overall_recall": 0.8784227437831796, "eval_runtime": 9.0651, "eval_samples_per_second": 938.328, "eval_steps_per_second": 29.343, "step": 14000 }, { "epoch": 2.87, "learning_rate": 3.564640665214809e-05, "loss": 0.484, "step": 14500 }, { "epoch": 2.97, "learning_rate": 3.515145515739457e-05, "loss": 0.4781, "step": 15000 }, { "epoch": 2.97, "eval_COMMENT": { "f1": 0.6936924742826204, "number": 6824, "precision": 0.6444919517102615, "recall": 0.7510257913247362 }, "eval_NAME": { "f1": 0.8070410729253981, "number": 8803, "precision": 0.7942146942366916, "recall": 0.8202885379984096 }, "eval_QTY": { "f1": 0.9859409799554566, "number": 7168, "precision": 0.98375, "recall": 0.9881417410714286 }, "eval_RANGE_END": { "f1": 0.6602870813397129, "number": 82, "precision": 0.5433070866141733, "recall": 0.8414634146341463 }, "eval_UNIT": { "f1": 0.9547484621218505, "number": 5755, "precision": 0.9268651832460733, "recall": 0.9843614248479583 }, "eval_loss": 0.48784932494163513, "eval_overall_accuracy": 0.8293345058978722, "eval_overall_f1": 0.8513236911105472, "eval_overall_precision": 0.825476495095627, "eval_overall_recall": 0.8788418552668343, "eval_runtime": 7.0221, "eval_samples_per_second": 1211.315, "eval_steps_per_second": 37.88, "step": 15000 }, { "epoch": 3.07, "learning_rate": 3.465650366264107e-05, "loss": 0.462, "step": 15500 }, { "epoch": 3.17, "learning_rate": 3.416155216788755e-05, "loss": 0.4599, "step": 16000 }, { "epoch": 3.17, "eval_COMMENT": { "f1": 0.6904161841836387, "number": 6824, "precision": 0.6450299096347206, "recall": 0.742672919109027 }, "eval_NAME": { "f1": 0.8082375265096551, "number": 8803, "precision": 0.7944048272078991, "recall": 0.8225604907417926 }, "eval_QTY": { "f1": 0.9857391304347826, "number": 7168, "precision": 0.983072013320383, "recall": 0.9884207589285714 }, "eval_RANGE_END": { "f1": 0.6636363636363637, "number": 82, "precision": 0.5289855072463768, "recall": 0.8902439024390244 }, "eval_UNIT": { "f1": 0.9555706291334577, "number": 5755, "precision": 0.9331015068719987, "recall": 0.9791485664639444 }, "eval_loss": 0.48515409231185913, "eval_overall_accuracy": 0.8305962873038025, "eval_overall_f1": 0.8510883569539567, "eval_overall_precision": 0.8269205428910265, "eval_overall_recall": 0.8767113718915899, "eval_runtime": 8.5334, "eval_samples_per_second": 996.787, "eval_steps_per_second": 31.172, "step": 16000 }, { "epoch": 3.27, "learning_rate": 3.3666600673134034e-05, "loss": 0.4548, "step": 16500 }, { "epoch": 3.37, "learning_rate": 3.317164917838052e-05, "loss": 0.4636, "step": 17000 }, { "epoch": 3.37, "eval_COMMENT": { "f1": 0.6982570806100218, "number": 6824, "precision": 0.6520854526958291, "recall": 0.7514654161781946 }, "eval_NAME": { "f1": 0.8060380709251347, "number": 8803, "precision": 0.7879774305555556, "recall": 0.8249460411223447 }, "eval_QTY": { "f1": 0.9860387580745988, "number": 7168, "precision": 0.9818785447503112, "recall": 0.990234375 }, "eval_RANGE_END": { "f1": 0.6198347107438017, "number": 82, "precision": 0.46875, "recall": 0.9146341463414634 }, "eval_UNIT": { "f1": 0.9546639751030365, "number": 5755, "precision": 0.9251711770459733, "recall": 0.9860990443092963 }, "eval_loss": 0.4889169931411743, "eval_overall_accuracy": 0.8302904009023648, "eval_overall_f1": 0.8521313412678315, "eval_overall_precision": 0.8246903898310624, "eval_overall_recall": 0.8814613020396759, "eval_runtime": 8.1836, "eval_samples_per_second": 1039.392, "eval_steps_per_second": 32.504, "step": 17000 }, { "epoch": 3.46, "learning_rate": 3.267669768362701e-05, "loss": 0.4638, "step": 17500 }, { "epoch": 3.56, "learning_rate": 3.218174618887349e-05, "loss": 0.4688, "step": 18000 }, { "epoch": 3.56, "eval_COMMENT": { "f1": 0.7010448264239973, "number": 6824, "precision": 0.6491074772188241, "recall": 0.7620164126611958 }, "eval_NAME": { "f1": 0.8081609837898266, "number": 8803, "precision": 0.7955320787938813, "recall": 0.8211973190957628 }, "eval_QTY": { "f1": 0.9860232250886588, "number": 7168, "precision": 0.9829474559822543, "recall": 0.9891183035714286 }, "eval_RANGE_END": { "f1": 0.6428571428571429, "number": 82, "precision": 0.5526315789473685, "recall": 0.7682926829268293 }, "eval_UNIT": { "f1": 0.9546716003700276, "number": 5755, "precision": 0.9250325945241199, "recall": 0.98627280625543 }, "eval_loss": 0.49013254046440125, "eval_overall_accuracy": 0.8319345403100923, "eval_overall_f1": 0.8534117209805214, "eval_overall_precision": 0.8264781911586663, "eval_overall_recall": 0.8821598211791003, "eval_runtime": 7.2351, "eval_samples_per_second": 1175.656, "eval_steps_per_second": 36.765, "step": 18000 }, { "epoch": 3.66, "learning_rate": 3.168679469411997e-05, "loss": 0.4648, "step": 18500 }, { "epoch": 3.76, "learning_rate": 3.119184319936647e-05, "loss": 0.4584, "step": 19000 }, { "epoch": 3.76, "eval_COMMENT": { "f1": 0.6999050332383666, "number": 6824, "precision": 0.65155342258146, "recall": 0.7560082063305978 }, "eval_NAME": { "f1": 0.8059467918622849, "number": 8803, "precision": 0.7932665859830564, "recall": 0.819038963989549 }, "eval_QTY": { "f1": 0.9858998402444954, "number": 7168, "precision": 0.981740213030848, "recall": 0.9900948660714286 }, "eval_RANGE_END": { "f1": 0.6636363636363637, "number": 82, "precision": 0.5289855072463768, "recall": 0.8902439024390244 }, "eval_UNIT": { "f1": 0.956632006086736, "number": 5755, "precision": 0.9315113598946329, "recall": 0.9831450912250217 }, "eval_loss": 0.48241040110588074, "eval_overall_accuracy": 0.8306727589041619, "eval_overall_f1": 0.8529790115098173, "eval_overall_precision": 0.8275420388859696, "eval_overall_recall": 0.8800293378038558, "eval_runtime": 9.4362, "eval_samples_per_second": 901.421, "eval_steps_per_second": 28.189, "step": 19000 }, { "epoch": 3.86, "learning_rate": 3.069689170461295e-05, "loss": 0.4584, "step": 19500 }, { "epoch": 3.96, "learning_rate": 3.0201940209859435e-05, "loss": 0.4701, "step": 20000 }, { "epoch": 3.96, "eval_COMMENT": { "f1": 0.7003649886371462, "number": 6824, "precision": 0.6606470053267507, "recall": 0.7451641266119577 }, "eval_NAME": { "f1": 0.8087504883084994, "number": 8803, "precision": 0.7948661693725319, "recall": 0.8231284789276383 }, "eval_QTY": { "f1": 0.98635477582846, "number": 7168, "precision": 0.9844357976653697, "recall": 0.98828125 }, "eval_RANGE_END": { "f1": 0.660377358490566, "number": 82, "precision": 0.5384615384615384, "recall": 0.8536585365853658 }, "eval_UNIT": { "f1": 0.956190798376184, "number": 5755, "precision": 0.9314549349151425, "recall": 0.9822762814943528 }, "eval_loss": 0.48059511184692383, "eval_overall_accuracy": 0.8328713174144952, "eval_overall_f1": 0.8544527532290959, "eval_overall_precision": 0.8321636652542372, "eval_overall_recall": 0.8779687063425538, "eval_runtime": 6.465, "eval_samples_per_second": 1315.706, "eval_steps_per_second": 41.145, "step": 20000 }, { "epoch": 4.06, "learning_rate": 2.970698871510592e-05, "loss": 0.4594, "step": 20500 }, { "epoch": 4.16, "learning_rate": 2.9212037220352405e-05, "loss": 0.4441, "step": 21000 }, { "epoch": 4.16, "eval_COMMENT": { "f1": 0.6940426407074792, "number": 6824, "precision": 0.6520674996779595, "recall": 0.7417936694021102 }, "eval_NAME": { "f1": 0.8061428889383486, "number": 8803, "precision": 0.7900534409423056, "recall": 0.8229012836533001 }, "eval_QTY": { "f1": 0.9851016429963798, "number": 7168, "precision": 0.9831851028349082, "recall": 0.9870256696428571 }, "eval_RANGE_END": { "f1": 0.6521739130434783, "number": 82, "precision": 0.5882352941176471, "recall": 0.7317073170731707 }, "eval_UNIT": { "f1": 0.9563670253325427, "number": 5755, "precision": 0.9332010582010583, "recall": 0.9807124239791486 }, "eval_loss": 0.4863247871398926, "eval_overall_accuracy": 0.8309977632056895, "eval_overall_f1": 0.851638092004753, "eval_overall_precision": 0.8284893321883876, "eval_overall_recall": 0.8761176306230791, "eval_runtime": 8.7832, "eval_samples_per_second": 968.441, "eval_steps_per_second": 30.285, "step": 21000 }, { "epoch": 4.26, "learning_rate": 2.871708572559889e-05, "loss": 0.4493, "step": 21500 }, { "epoch": 4.36, "learning_rate": 2.8222134230845377e-05, "loss": 0.4375, "step": 22000 }, { "epoch": 4.36, "eval_COMMENT": { "f1": 0.699911438108863, "number": 6824, "precision": 0.6539783577339274, "recall": 0.7527842907385698 }, "eval_NAME": { "f1": 0.8084585420144685, "number": 8803, "precision": 0.7924075488164066, "recall": 0.8251732363966829 }, "eval_QTY": { "f1": 0.985897881208753, "number": 7168, "precision": 0.9818735298187353, "recall": 0.9899553571428571 }, "eval_RANGE_END": { "f1": 0.6519823788546256, "number": 82, "precision": 0.5103448275862069, "recall": 0.9024390243902439 }, "eval_UNIT": { "f1": 0.9565510841137266, "number": 5755, "precision": 0.9296490652673007, "recall": 0.9850564726324935 }, "eval_loss": 0.4853152334690094, "eval_overall_accuracy": 0.8326036668132373, "eval_overall_f1": 0.8537987957513024, "eval_overall_precision": 0.8277581004853732, "eval_overall_recall": 0.8815311539536184, "eval_runtime": 7.6535, "eval_samples_per_second": 1111.391, "eval_steps_per_second": 34.755, "step": 22000 }, { "epoch": 4.45, "learning_rate": 2.7727182736091867e-05, "loss": 0.4541, "step": 22500 }, { "epoch": 4.55, "learning_rate": 2.7232231241338353e-05, "loss": 0.4566, "step": 23000 }, { "epoch": 4.55, "eval_COMMENT": { "f1": 0.6999176728869374, "number": 6824, "precision": 0.6580237358101135, "recall": 0.7475087924970691 }, "eval_NAME": { "f1": 0.8100474462740721, "number": 8803, "precision": 0.7963125548726954, "recall": 0.8242644552993298 }, "eval_QTY": { "f1": 0.9862956521739129, "number": 7168, "precision": 0.9836270292770917, "recall": 0.9889787946428571 }, "eval_RANGE_END": { "f1": 0.6979166666666666, "number": 82, "precision": 0.6090909090909091, "recall": 0.8170731707317073 }, "eval_UNIT": { "f1": 0.9566618607412434, "number": 5755, "precision": 0.9343936381709742, "recall": 0.9800173761946134 }, "eval_loss": 0.48071911931037903, "eval_overall_accuracy": 0.8325271952128778, "eval_overall_f1": 0.8548318578055702, "eval_overall_precision": 0.8324122182877188, "eval_overall_recall": 0.8784925956971221, "eval_runtime": 7.014, "eval_samples_per_second": 1212.724, "eval_steps_per_second": 37.924, "step": 23000 }, { "epoch": 4.65, "learning_rate": 2.6737279746584836e-05, "loss": 0.4448, "step": 23500 }, { "epoch": 4.75, "learning_rate": 2.6242328251831323e-05, "loss": 0.4546, "step": 24000 }, { "epoch": 4.75, "eval_COMMENT": { "f1": 0.7049292873815736, "number": 6824, "precision": 0.6631361405321622, "recall": 0.7523446658851114 }, "eval_NAME": { "f1": 0.8063276332646355, "number": 8803, "precision": 0.7910382513661203, "recall": 0.8222196978302851 }, "eval_QTY": { "f1": 0.9859881491808993, "number": 7168, "precision": 0.985369931726348, "recall": 0.9866071428571429 }, "eval_RANGE_END": { "f1": 0.6783625730994152, "number": 82, "precision": 0.651685393258427, "recall": 0.7073170731707317 }, "eval_UNIT": { "f1": 0.9558001693480102, "number": 5755, "precision": 0.9321222130470685, "recall": 0.9807124239791486 }, "eval_loss": 0.47966957092285156, "eval_overall_accuracy": 0.832374252012159, "eval_overall_f1": 0.8546520520010196, "eval_overall_precision": 0.8322907357759904, "eval_overall_recall": 0.8782481139983236, "eval_runtime": 9.0205, "eval_samples_per_second": 942.965, "eval_steps_per_second": 29.488, "step": 24000 }, { "epoch": 4.85, "learning_rate": 2.5747376757077806e-05, "loss": 0.4498, "step": 24500 }, { "epoch": 4.95, "learning_rate": 2.5252425262324292e-05, "loss": 0.4462, "step": 25000 }, { "epoch": 4.95, "eval_COMMENT": { "f1": 0.6998077451249656, "number": 6824, "precision": 0.6583979328165375, "recall": 0.7467760844079718 }, "eval_NAME": { "f1": 0.8070350643586329, "number": 8803, "precision": 0.7887430864331417, "recall": 0.8261956151312053 }, "eval_QTY": { "f1": 0.9860329372524493, "number": 7168, "precision": 0.9822788315104527, "recall": 0.9898158482142857 }, "eval_RANGE_END": { "f1": 0.6571428571428573, "number": 82, "precision": 0.5390625, "recall": 0.8414634146341463 }, "eval_UNIT": { "f1": 0.9555161644298135, "number": 5755, "precision": 0.9290873276428102, "recall": 0.9834926151172894 }, "eval_loss": 0.48545241355895996, "eval_overall_accuracy": 0.8324889594126981, "eval_overall_f1": 0.8534792330103665, "eval_overall_precision": 0.8286080778844889, "eval_overall_recall": 0.879889633975971, "eval_runtime": 6.8304, "eval_samples_per_second": 1245.307, "eval_steps_per_second": 38.943, "step": 25000 }, { "epoch": 5.05, "learning_rate": 2.4757473767570778e-05, "loss": 0.4373, "step": 25500 }, { "epoch": 5.15, "learning_rate": 2.4262522272817265e-05, "loss": 0.433, "step": 26000 }, { "epoch": 5.15, "eval_COMMENT": { "f1": 0.7011549238023645, "number": 6824, "precision": 0.656934306569343, "recall": 0.7517584994138335 }, "eval_NAME": { "f1": 0.805529234713784, "number": 8803, "precision": 0.7907638432917488, "recall": 0.8208565261842554 }, "eval_QTY": { "f1": 0.9860607750209088, "number": 7168, "precision": 0.9852367688022284, "recall": 0.9868861607142857 }, "eval_RANGE_END": { "f1": 0.6607929515418502, "number": 82, "precision": 0.5172413793103449, "recall": 0.9146341463414634 }, "eval_UNIT": { "f1": 0.9570749915167968, "number": 5755, "precision": 0.935024034477043, "recall": 0.9801911381407472 }, "eval_loss": 0.4933657944202423, "eval_overall_accuracy": 0.8298315713002083, "eval_overall_f1": 0.8533179496750768, "eval_overall_precision": 0.8297640653357532, "eval_overall_recall": 0.8782481139983236, "eval_runtime": 8.2218, "eval_samples_per_second": 1034.568, "eval_steps_per_second": 32.353, "step": 26000 }, { "epoch": 5.25, "learning_rate": 2.376757077806375e-05, "loss": 0.4353, "step": 26500 }, { "epoch": 5.35, "learning_rate": 2.3272619283310237e-05, "loss": 0.442, "step": 27000 }, { "epoch": 5.35, "eval_COMMENT": { "f1": 0.7040608447643624, "number": 6824, "precision": 0.6560364464692483, "recall": 0.7596717467760844 }, "eval_NAME": { "f1": 0.8098804068402817, "number": 8803, "precision": 0.7970520294797052, "recall": 0.8231284789276383 }, "eval_QTY": { "f1": 0.9860037601838312, "number": 7168, "precision": 0.9842902822188239, "recall": 0.9877232142857143 }, "eval_RANGE_END": { "f1": 0.6728110599078342, "number": 82, "precision": 0.5407407407407407, "recall": 0.8902439024390244 }, "eval_UNIT": { "f1": 0.956294296321296, "number": 5755, "precision": 0.9294735115630638, "recall": 0.9847089487402259 }, "eval_loss": 0.48500868678092957, "eval_overall_accuracy": 0.8332727933163822, "eval_overall_f1": 0.855207451312447, "eval_overall_precision": 0.8301005983299362, "eval_overall_recall": 0.8818804135233306, "eval_runtime": 8.8139, "eval_samples_per_second": 965.065, "eval_steps_per_second": 30.18, "step": 27000 }, { "epoch": 5.44, "learning_rate": 2.2777667788556724e-05, "loss": 0.4321, "step": 27500 }, { "epoch": 5.54, "learning_rate": 2.2282716293803206e-05, "loss": 0.4348, "step": 28000 }, { "epoch": 5.54, "eval_COMMENT": { "f1": 0.7028109676082602, "number": 6824, "precision": 0.6646636185499674, "recall": 0.7456037514654161 }, "eval_NAME": { "f1": 0.8111872527717422, "number": 8803, "precision": 0.7959763831183031, "recall": 0.8269907985913894 }, "eval_QTY": { "f1": 0.9861033907726514, "number": 7168, "precision": 0.982281284606866, "recall": 0.9899553571428571 }, "eval_RANGE_END": { "f1": 0.6995073891625616, "number": 82, "precision": 0.5867768595041323, "recall": 0.8658536585365854 }, "eval_UNIT": { "f1": 0.9556492411467116, "number": 5755, "precision": 0.9282555282555283, "recall": 0.9847089487402259 }, "eval_loss": 0.4908902645111084, "eval_overall_accuracy": 0.8319536582101822, "eval_overall_f1": 0.8560025813902146, "eval_overall_precision": 0.8330964265644111, "eval_overall_recall": 0.8802039675887119, "eval_runtime": 6.4986, "eval_samples_per_second": 1308.901, "eval_steps_per_second": 40.932, "step": 28000 }, { "epoch": 5.64, "learning_rate": 2.1787764799049696e-05, "loss": 0.4335, "step": 28500 }, { "epoch": 5.74, "learning_rate": 2.129281330429618e-05, "loss": 0.4324, "step": 29000 }, { "epoch": 5.74, "eval_COMMENT": { "f1": 0.7066996835878389, "number": 6824, "precision": 0.6659320715582059, "recall": 0.7527842907385698 }, "eval_NAME": { "f1": 0.8106571540047935, "number": 8803, "precision": 0.7957977675640184, "recall": 0.8260820174940361 }, "eval_QTY": { "f1": 0.9864479810966712, "number": 7168, "precision": 0.9828278631768453, "recall": 0.9900948660714286 }, "eval_RANGE_END": { "f1": 0.7035175879396985, "number": 82, "precision": 0.5982905982905983, "recall": 0.8536585365853658 }, "eval_UNIT": { "f1": 0.9554875976894326, "number": 5755, "precision": 0.9346850589995014, "recall": 0.9772371850564726 }, "eval_loss": 0.4821859300136566, "eval_overall_accuracy": 0.8341139809203357, "eval_overall_f1": 0.8565747208484169, "eval_overall_precision": 0.8342437183434304, "eval_overall_recall": 0.8801341156747695, "eval_runtime": 8.6529, "eval_samples_per_second": 983.028, "eval_steps_per_second": 30.741, "step": 29000 }, { "epoch": 5.84, "learning_rate": 2.0797861809542665e-05, "loss": 0.4433, "step": 29500 }, { "epoch": 5.94, "learning_rate": 2.0302910314789152e-05, "loss": 0.4375, "step": 30000 }, { "epoch": 5.94, "eval_COMMENT": { "f1": 0.70434363598816, "number": 6824, "precision": 0.6641568220173958, "recall": 0.7497069167643611 }, "eval_NAME": { "f1": 0.8090376569037656, "number": 8803, "precision": 0.7948914711686034, "recall": 0.823696467113484 }, "eval_QTY": { "f1": 0.9860154456272177, "number": 7168, "precision": 0.9834836918806384, "recall": 0.9885602678571429 }, "eval_RANGE_END": { "f1": 0.6785714285714285, "number": 82, "precision": 0.5352112676056338, "recall": 0.926829268292683 }, "eval_UNIT": { "f1": 0.9560281284419215, "number": 5755, "precision": 0.9328703703703703, "recall": 0.9803649000868809 }, "eval_loss": 0.48426133394241333, "eval_overall_accuracy": 0.8327948458141358, "eval_overall_f1": 0.8553999864065792, "eval_overall_precision": 0.8329252150893448, "eval_overall_recall": 0.8791212629226041, "eval_runtime": 7.6434, "eval_samples_per_second": 1112.862, "eval_steps_per_second": 34.801, "step": 30000 }, { "epoch": 6.04, "learning_rate": 1.9807958820035638e-05, "loss": 0.4331, "step": 30500 }, { "epoch": 6.14, "learning_rate": 1.9313007325282124e-05, "loss": 0.4226, "step": 31000 }, { "epoch": 6.14, "eval_COMMENT": { "f1": 0.7090600610790635, "number": 6824, "precision": 0.6603463531791177, "recall": 0.7655334114888629 }, "eval_NAME": { "f1": 0.806347438752784, "number": 8803, "precision": 0.7907611663208475, "recall": 0.8225604907417926 }, "eval_QTY": { "f1": 0.9856705620478576, "number": 7168, "precision": 0.9829356270810211, "recall": 0.9884207589285714 }, "eval_RANGE_END": { "f1": 0.669603524229075, "number": 82, "precision": 0.5241379310344828, "recall": 0.926829268292683 }, "eval_UNIT": { "f1": 0.9566317921025986, "number": 5755, "precision": 0.9298015417418403, "recall": 0.9850564726324935 }, "eval_loss": 0.48777666687965393, "eval_overall_accuracy": 0.8335786797178198, "eval_overall_f1": 0.8552831783601016, "eval_overall_precision": 0.8288551019070712, "eval_overall_recall": 0.8834520815870355, "eval_runtime": 8.452, "eval_samples_per_second": 1006.385, "eval_steps_per_second": 31.472, "step": 31000 }, { "epoch": 6.24, "learning_rate": 1.8818055830528607e-05, "loss": 0.4237, "step": 31500 }, { "epoch": 6.34, "learning_rate": 1.8323104335775097e-05, "loss": 0.4305, "step": 32000 }, { "epoch": 6.34, "eval_COMMENT": { "f1": 0.7089205043489988, "number": 6824, "precision": 0.6565505182964906, "recall": 0.770369284876905 }, "eval_NAME": { "f1": 0.8125769273805529, "number": 8803, "precision": 0.8005732554293904, "recall": 0.8249460411223447 }, "eval_QTY": { "f1": 0.9859881491808993, "number": 7168, "precision": 0.985369931726348, "recall": 0.9866071428571429 }, "eval_RANGE_END": { "f1": 0.6634146341463415, "number": 82, "precision": 0.5528455284552846, "recall": 0.8292682926829268 }, "eval_UNIT": { "f1": 0.956661316211878, "number": 5755, "precision": 0.9309437684972048, "recall": 0.9838401390095569 }, "eval_loss": 0.4837185740470886, "eval_overall_accuracy": 0.834209570420785, "eval_overall_f1": 0.8570026399512625, "eval_overall_precision": 0.8312869336835194, "eval_overall_recall": 0.8843601564682873, "eval_runtime": 9.3183, "eval_samples_per_second": 912.826, "eval_steps_per_second": 28.546, "step": 32000 }, { "epoch": 6.43, "learning_rate": 1.782815284102158e-05, "loss": 0.423, "step": 32500 }, { "epoch": 6.53, "learning_rate": 1.7333201346268066e-05, "loss": 0.4142, "step": 33000 }, { "epoch": 6.53, "eval_COMMENT": { "f1": 0.6982502247734974, "number": 6824, "precision": 0.6611656843483955, "recall": 0.7397420867526378 }, "eval_NAME": { "f1": 0.8101717599821548, "number": 8803, "precision": 0.7957059918939643, "recall": 0.8251732363966829 }, "eval_QTY": { "f1": 0.9850003456141564, "number": 7168, "precision": 0.9761611179613646, "recall": 0.9940011160714286 }, "eval_RANGE_END": { "f1": 0.669683257918552, "number": 82, "precision": 0.5323741007194245, "recall": 0.9024390243902439 }, "eval_UNIT": { "f1": 0.9568381855111713, "number": 5755, "precision": 0.9326843755155915, "recall": 0.9822762814943528 }, "eval_loss": 0.48620671033859253, "eval_overall_accuracy": 0.833158085915843, "eval_overall_f1": 0.8545377366499703, "eval_overall_precision": 0.831510425271784, "eval_overall_recall": 0.8788767812238055, "eval_runtime": 6.5738, "eval_samples_per_second": 1293.917, "eval_steps_per_second": 40.463, "step": 33000 }, { "epoch": 6.63, "learning_rate": 1.6838249851514553e-05, "loss": 0.4247, "step": 33500 }, { "epoch": 6.73, "learning_rate": 1.634329835676104e-05, "loss": 0.4327, "step": 34000 }, { "epoch": 6.73, "eval_COMMENT": { "f1": 0.7027439024390244, "number": 6824, "precision": 0.6665352260778128, "recall": 0.7431125439624854 }, "eval_NAME": { "f1": 0.8050692012673003, "number": 8803, "precision": 0.7882020026121027, "recall": 0.8226740883789617 }, "eval_QTY": { "f1": 0.9866000138859959, "number": 7168, "precision": 0.9820317899101589, "recall": 0.9912109375 }, "eval_RANGE_END": { "f1": 0.6930693069306931, "number": 82, "precision": 0.5833333333333334, "recall": 0.8536585365853658 }, "eval_UNIT": { "f1": 0.9562352145995269, "number": 5755, "precision": 0.9306035191580332, "recall": 0.9833188531711555 }, "eval_loss": 0.4834836721420288, "eval_overall_accuracy": 0.8330242606152141, "eval_overall_f1": 0.8544101658059254, "eval_overall_precision": 0.8318007409367557, "eval_overall_recall": 0.8782830399552948, "eval_runtime": 8.5374, "eval_samples_per_second": 996.324, "eval_steps_per_second": 31.157, "step": 34000 }, { "epoch": 6.83, "learning_rate": 1.5848346862007525e-05, "loss": 0.4289, "step": 34500 }, { "epoch": 6.93, "learning_rate": 1.535339536725401e-05, "loss": 0.4299, "step": 35000 }, { "epoch": 6.93, "eval_COMMENT": { "f1": 0.706956462492335, "number": 6824, "precision": 0.660639246147969, "recall": 0.7602579132473622 }, "eval_NAME": { "f1": 0.8096777790166128, "number": 8803, "precision": 0.7949644225506295, "recall": 0.8249460411223447 }, "eval_QTY": { "f1": 0.9858723641171967, "number": 7168, "precision": 0.9836133870295792, "recall": 0.9881417410714286 }, "eval_RANGE_END": { "f1": 0.6757990867579909, "number": 82, "precision": 0.5401459854014599, "recall": 0.9024390243902439 }, "eval_UNIT": { "f1": 0.9569053274516457, "number": 5755, "precision": 0.9348582794629537, "recall": 0.9800173761946134 }, "eval_loss": 0.4832661747932434, "eval_overall_accuracy": 0.8333301470166516, "eval_overall_f1": 0.8559610788086318, "eval_overall_precision": 0.8316150070819197, "eval_overall_recall": 0.8817756356524169, "eval_runtime": 7.9218, "eval_samples_per_second": 1073.752, "eval_steps_per_second": 33.578, "step": 35000 }, { "epoch": 7.03, "learning_rate": 1.4858443872500496e-05, "loss": 0.4181, "step": 35500 }, { "epoch": 7.13, "learning_rate": 1.4363492377746981e-05, "loss": 0.4165, "step": 36000 }, { "epoch": 7.13, "eval_COMMENT": { "f1": 0.7096729965037363, "number": 6824, "precision": 0.6667525441195414, "recall": 0.7584994138335287 }, "eval_NAME": { "f1": 0.8082838265504414, "number": 8803, "precision": 0.7905082536924414, "recall": 0.8268772009542201 }, "eval_QTY": { "f1": 0.9864479810966712, "number": 7168, "precision": 0.9828278631768453, "recall": 0.9900948660714286 }, "eval_RANGE_END": { "f1": 0.6542056074766355, "number": 82, "precision": 0.5303030303030303, "recall": 0.8536585365853658 }, "eval_UNIT": { "f1": 0.9563968227142133, "number": 5755, "precision": 0.930909689093601, "recall": 0.9833188531711555 }, "eval_loss": 0.48711806535720825, "eval_overall_accuracy": 0.8347831074234806, "eval_overall_f1": 0.8564749724739562, "eval_overall_precision": 0.8315297832450745, "eval_overall_recall": 0.8829631181894384, "eval_runtime": 8.4247, "eval_samples_per_second": 1009.646, "eval_steps_per_second": 31.574, "step": 36000 }, { "epoch": 7.23, "learning_rate": 1.3868540882993467e-05, "loss": 0.4199, "step": 36500 }, { "epoch": 7.33, "learning_rate": 1.3373589388239954e-05, "loss": 0.4213, "step": 37000 }, { "epoch": 7.33, "eval_COMMENT": { "f1": 0.7081497797356828, "number": 6824, "precision": 0.667705088265836, "recall": 0.753810082063306 }, "eval_NAME": { "f1": 0.8101195440644984, "number": 8803, "precision": 0.7934001306904814, "recall": 0.827558786777235 }, "eval_QTY": { "f1": 0.9860821155184412, "number": 7168, "precision": 0.983754512635379, "recall": 0.9884207589285714 }, "eval_RANGE_END": { "f1": 0.663594470046083, "number": 82, "precision": 0.5333333333333333, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.9559865092748736, "number": 5755, "precision": 0.9285831285831286, "recall": 0.9850564726324935 }, "eval_loss": 0.4881742596626282, "eval_overall_accuracy": 0.8332727933163822, "eval_overall_f1": 0.8566824966078698, "eval_overall_precision": 0.832728831442891, "eval_overall_recall": 0.8820550433081866, "eval_runtime": 8.1883, "eval_samples_per_second": 1038.805, "eval_steps_per_second": 32.486, "step": 37000 }, { "epoch": 7.42, "learning_rate": 1.287863789348644e-05, "loss": 0.4196, "step": 37500 }, { "epoch": 7.52, "learning_rate": 1.2383686398732925e-05, "loss": 0.4122, "step": 38000 }, { "epoch": 7.52, "eval_COMMENT": { "f1": 0.7068037866920929, "number": 6824, "precision": 0.660262119862578, "recall": 0.7604044548651817 }, "eval_NAME": { "f1": 0.8086100450525612, "number": 8803, "precision": 0.7921752397558849, "recall": 0.8257412245825286 }, "eval_QTY": { "f1": 0.9867148918411351, "number": 7168, "precision": 0.9839090026355944, "recall": 0.9895368303571429 }, "eval_RANGE_END": { "f1": 0.6766169154228855, "number": 82, "precision": 0.5714285714285714, "recall": 0.8292682926829268 }, "eval_UNIT": { "f1": 0.9566393373341222, "number": 5755, "precision": 0.9313693219223174, "recall": 0.9833188531711555 }, "eval_loss": 0.4913991689682007, "eval_overall_accuracy": 0.8336933871183589, "eval_overall_f1": 0.8558514330212795, "eval_overall_precision": 0.8304477808075167, "eval_overall_recall": 0.8828583403185247, "eval_runtime": 6.6374, "eval_samples_per_second": 1281.524, "eval_steps_per_second": 40.076, "step": 38000 }, { "epoch": 7.62, "learning_rate": 1.1888734903979411e-05, "loss": 0.4196, "step": 38500 }, { "epoch": 7.72, "learning_rate": 1.1393783409225896e-05, "loss": 0.4034, "step": 39000 }, { "epoch": 7.72, "eval_COMMENT": { "f1": 0.7121398334016114, "number": 6824, "precision": 0.6667092815136794, "recall": 0.7642145369284877 }, "eval_NAME": { "f1": 0.8117607676857844, "number": 8803, "precision": 0.7976099111939481, "recall": 0.8264228104055436 }, "eval_QTY": { "f1": 0.9866388308977034, "number": 7168, "precision": 0.984309913912802, "recall": 0.9889787946428571 }, "eval_RANGE_END": { "f1": 0.6859903381642511, "number": 82, "precision": 0.568, "recall": 0.8658536585365854 }, "eval_UNIT": { "f1": 0.9563307711799983, "number": 5755, "precision": 0.9304733727810651, "recall": 0.9836663770634231 }, "eval_loss": 0.4883587956428528, "eval_overall_accuracy": 0.8344007494216835, "eval_overall_f1": 0.8582036415420609, "eval_overall_precision": 0.8338604467286025, "eval_overall_recall": 0.884010896898575, "eval_runtime": 9.4932, "eval_samples_per_second": 896.008, "eval_steps_per_second": 28.02, "step": 39000 }, { "epoch": 7.82, "learning_rate": 1.0898831914472382e-05, "loss": 0.4145, "step": 39500 }, { "epoch": 7.92, "learning_rate": 1.0403880419718868e-05, "loss": 0.4209, "step": 40000 }, { "epoch": 7.92, "eval_COMMENT": { "f1": 0.706067877956805, "number": 6824, "precision": 0.6634454322896534, "recall": 0.7545427901524033 }, "eval_NAME": { "f1": 0.81135449262405, "number": 8803, "precision": 0.7984163642362256, "recall": 0.8247188458480064 }, "eval_QTY": { "f1": 0.9865127919911012, "number": 7168, "precision": 0.9832317073170732, "recall": 0.9898158482142857 }, "eval_RANGE_END": { "f1": 0.6697674418604651, "number": 82, "precision": 0.5413533834586466, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.9571609632446134, "number": 5755, "precision": 0.9315789473684211, "recall": 0.9841876629018245 }, "eval_loss": 0.49103492498397827, "eval_overall_accuracy": 0.8326992563136865, "eval_overall_f1": 0.8568276330306375, "eval_overall_precision": 0.8334709242809497, "eval_overall_recall": 0.8815311539536184, "eval_runtime": 6.7729, "eval_samples_per_second": 1255.88, "eval_steps_per_second": 39.274, "step": 40000 }, { "epoch": 8.02, "learning_rate": 9.908928924965353e-06, "loss": 0.4146, "step": 40500 }, { "epoch": 8.12, "learning_rate": 9.41397743021184e-06, "loss": 0.4078, "step": 41000 }, { "epoch": 8.12, "eval_COMMENT": { "f1": 0.7065016749846176, "number": 6824, "precision": 0.6621812123542228, "recall": 0.7571805392731535 }, "eval_NAME": { "f1": 0.8095132004010249, "number": 8803, "precision": 0.7941208611080757, "recall": 0.8255140293081904 }, "eval_QTY": { "f1": 0.9861565217391305, "number": 7168, "precision": 0.9834882752879145, "recall": 0.9888392857142857 }, "eval_RANGE_END": { "f1": 0.6824644549763034, "number": 82, "precision": 0.5581395348837209, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.9566246723598546, "number": 5755, "precision": 0.9316534914361001, "recall": 0.982971329278888 }, "eval_loss": 0.49227482080459595, "eval_overall_accuracy": 0.8334830902173705, "eval_overall_f1": 0.856053157948266, "eval_overall_precision": 0.8316645807259074, "eval_overall_recall": 0.8819153394803018, "eval_runtime": 8.7342, "eval_samples_per_second": 973.868, "eval_steps_per_second": 30.455, "step": 41000 }, { "epoch": 8.22, "learning_rate": 8.919025935458326e-06, "loss": 0.4008, "step": 41500 }, { "epoch": 8.32, "learning_rate": 8.424074440704812e-06, "loss": 0.4098, "step": 42000 }, { "epoch": 8.32, "eval_COMMENT": { "f1": 0.7069577080491133, "number": 6824, "precision": 0.6613067891781521, "recall": 0.7593786635404455 }, "eval_NAME": { "f1": 0.8097364895042429, "number": 8803, "precision": 0.7961356899769458, "recall": 0.8238100647506532 }, "eval_QTY": { "f1": 0.9860801781737194, "number": 7168, "precision": 0.9838888888888889, "recall": 0.98828125 }, "eval_RANGE_END": { "f1": 0.6889952153110047, "number": 82, "precision": 0.5669291338582677, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.9567859554355165, "number": 5755, "precision": 0.9302478253733792, "recall": 0.9848827106863597 }, "eval_loss": 0.4903165102005005, "eval_overall_accuracy": 0.8344198673217734, "eval_overall_f1": 0.8562469278098886, "eval_overall_precision": 0.831812942532521, "eval_overall_recall": 0.8821598211791003, "eval_runtime": 7.5483, "eval_samples_per_second": 1126.874, "eval_steps_per_second": 35.24, "step": 42000 }, { "epoch": 8.41, "learning_rate": 7.929122945951298e-06, "loss": 0.4092, "step": 42500 }, { "epoch": 8.51, "learning_rate": 7.434171451197783e-06, "loss": 0.4009, "step": 43000 }, { "epoch": 8.51, "eval_COMMENT": { "f1": 0.7111504907306434, "number": 6824, "precision": 0.6647553516819572, "recall": 0.7645076201641267 }, "eval_NAME": { "f1": 0.8115311698449873, "number": 8803, "precision": 0.7969554265688314, "recall": 0.8266500056798819 }, "eval_QTY": { "f1": 0.9865015307542444, "number": 7168, "precision": 0.9840366463076069, "recall": 0.9889787946428571 }, "eval_RANGE_END": { "f1": 0.6666666666666667, "number": 82, "precision": 0.5689655172413793, "recall": 0.8048780487804879 }, "eval_UNIT": { "f1": 0.9566759564225994, "number": 5755, "precision": 0.9306605323693723, "recall": 0.9841876629018245 }, "eval_loss": 0.4936346113681793, "eval_overall_accuracy": 0.8347639895233907, "eval_overall_f1": 0.8578206279546571, "eval_overall_precision": 0.8330755306894849, "eval_overall_recall": 0.8840807488125174, "eval_runtime": 7.8937, "eval_samples_per_second": 1077.574, "eval_steps_per_second": 33.698, "step": 43000 }, { "epoch": 8.61, "learning_rate": 6.939219956444269e-06, "loss": 0.4103, "step": 43500 }, { "epoch": 8.71, "learning_rate": 6.444268461690754e-06, "loss": 0.41, "step": 44000 }, { "epoch": 8.71, "eval_COMMENT": { "f1": 0.7139846352573254, "number": 6824, "precision": 0.6659480025364616, "recall": 0.7694900351699883 }, "eval_NAME": { "f1": 0.8144254962258876, "number": 8803, "precision": 0.8019158775600088, "recall": 0.8273315915028967 }, "eval_QTY": { "f1": 0.9864385562278323, "number": 7168, "precision": 0.9834974344751075, "recall": 0.9893973214285714 }, "eval_RANGE_END": { "f1": 0.6728110599078342, "number": 82, "precision": 0.5407407407407407, "recall": 0.8902439024390244 }, "eval_UNIT": { "f1": 0.9569402228976698, "number": 5755, "precision": 0.9306946953522746, "recall": 0.9847089487402259 }, "eval_loss": 0.4894082546234131, "eval_overall_accuracy": 0.835911063528782, "eval_overall_f1": 0.8593691770843921, "eval_overall_precision": 0.8343530030919019, "eval_overall_recall": 0.8859318245319922, "eval_runtime": 9.03, "eval_samples_per_second": 941.975, "eval_steps_per_second": 29.457, "step": 44000 }, { "epoch": 8.81, "learning_rate": 5.949316966937241e-06, "loss": 0.4063, "step": 44500 }, { "epoch": 8.91, "learning_rate": 5.4543654721837265e-06, "loss": 0.4177, "step": 45000 }, { "epoch": 8.91, "eval_COMMENT": { "f1": 0.7089166264395559, "number": 6824, "precision": 0.6695323694151362, "recall": 0.7532239155920282 }, "eval_NAME": { "f1": 0.810614930032893, "number": 8803, "precision": 0.7959273045763083, "recall": 0.8258548222196979 }, "eval_QTY": { "f1": 0.9862956521739129, "number": 7168, "precision": 0.9836270292770917, "recall": 0.9889787946428571 }, "eval_RANGE_END": { "f1": 0.6926829268292684, "number": 82, "precision": 0.5772357723577236, "recall": 0.8658536585365854 }, "eval_UNIT": { "f1": 0.9571682014023823, "number": 5755, "precision": 0.9314370272936534, "recall": 0.9843614248479583 }, "eval_loss": 0.4893593490123749, "eval_overall_accuracy": 0.8339036840193473, "eval_overall_f1": 0.8575312207968737, "eval_overall_precision": 0.8349601297025444, "eval_overall_recall": 0.8813565241687622, "eval_runtime": 7.9862, "eval_samples_per_second": 1065.084, "eval_steps_per_second": 33.307, "step": 45000 }, { "epoch": 9.01, "learning_rate": 4.959413977430212e-06, "loss": 0.4109, "step": 45500 }, { "epoch": 9.11, "learning_rate": 4.4644624826766974e-06, "loss": 0.399, "step": 46000 }, { "epoch": 9.11, "eval_COMMENT": { "f1": 0.7139457401237506, "number": 6824, "precision": 0.6659901052898642, "recall": 0.7693434935521688 }, "eval_NAME": { "f1": 0.8108047897521581, "number": 8803, "precision": 0.7953452797202797, "recall": 0.8268772009542201 }, "eval_QTY": { "f1": 0.9858156028368794, "number": 7168, "precision": 0.9826725810923205, "recall": 0.9889787946428571 }, "eval_RANGE_END": { "f1": 0.6790697674418604, "number": 82, "precision": 0.5488721804511278, "recall": 0.8902439024390244 }, "eval_UNIT": { "f1": 0.9572476600050595, "number": 5755, "precision": 0.9298820445609436, "recall": 0.98627280625543 }, "eval_loss": 0.49086296558380127, "eval_overall_accuracy": 0.8356434129275241, "eval_overall_f1": 0.8581819412023411, "eval_overall_precision": 0.832086859542085, "eval_overall_recall": 0.8859667504889635, "eval_runtime": 9.6791, "eval_samples_per_second": 878.799, "eval_steps_per_second": 27.482, "step": 46000 }, { "epoch": 9.21, "learning_rate": 3.969510987923184e-06, "loss": 0.4028, "step": 46500 }, { "epoch": 9.31, "learning_rate": 3.4745594931696697e-06, "loss": 0.4066, "step": 47000 }, { "epoch": 9.31, "eval_COMMENT": { "f1": 0.7096332785987959, "number": 6824, "precision": 0.6655544147843943, "recall": 0.7599648300117233 }, "eval_NAME": { "f1": 0.8115457483561797, "number": 8803, "precision": 0.7964563053702286, "recall": 0.8272179938657276 }, "eval_QTY": { "f1": 0.9858900396191006, "number": 7168, "precision": 0.9824075356697604, "recall": 0.9893973214285714 }, "eval_RANGE_END": { "f1": 0.6792452830188679, "number": 82, "precision": 0.5538461538461539, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.9569838056680162, "number": 5755, "precision": 0.9298475659727914, "recall": 0.9857515204170286 }, "eval_loss": 0.4892035126686096, "eval_overall_accuracy": 0.8339610377196168, "eval_overall_f1": 0.8575495196299372, "eval_overall_precision": 0.8328122428830015, "eval_overall_recall": 0.8838013411567477, "eval_runtime": 7.0853, "eval_samples_per_second": 1200.507, "eval_steps_per_second": 37.542, "step": 47000 }, { "epoch": 9.4, "learning_rate": 2.979607998416155e-06, "loss": 0.4045, "step": 47500 }, { "epoch": 9.5, "learning_rate": 2.484656503662641e-06, "loss": 0.397, "step": 48000 }, { "epoch": 9.5, "eval_COMMENT": { "f1": 0.7098901098901098, "number": 6824, "precision": 0.6680455015511892, "recall": 0.757327080890973 }, "eval_NAME": { "f1": 0.8101547711836098, "number": 8803, "precision": 0.7944098700731521, "recall": 0.8265364080427127 }, "eval_QTY": { "f1": 0.9862327909887358, "number": 7168, "precision": 0.9830884391461048, "recall": 0.9893973214285714 }, "eval_RANGE_END": { "f1": 0.6857142857142857, "number": 82, "precision": 0.5625, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.9567713610266801, "number": 5755, "precision": 0.9305304647725406, "recall": 0.9845351867940921 }, "eval_loss": 0.491767019033432, "eval_overall_accuracy": 0.8346875179230313, "eval_overall_f1": 0.8573560839919944, "eval_overall_precision": 0.8334102750115413, "eval_overall_recall": 0.8827186364906399, "eval_runtime": 8.8099, "eval_samples_per_second": 965.502, "eval_steps_per_second": 30.193, "step": 48000 }, { "epoch": 9.6, "learning_rate": 1.989705008909127e-06, "loss": 0.4023, "step": 48500 }, { "epoch": 9.7, "learning_rate": 1.4947535141556129e-06, "loss": 0.3987, "step": 49000 }, { "epoch": 9.7, "eval_COMMENT": { "f1": 0.7078929306794783, "number": 6824, "precision": 0.6657629744384198, "recall": 0.7557151230949589 }, "eval_NAME": { "f1": 0.8108077994428969, "number": 8803, "precision": 0.7955613862468569, "recall": 0.8266500056798819 }, "eval_QTY": { "f1": 0.9860232250886588, "number": 7168, "precision": 0.9829474559822543, "recall": 0.9891183035714286 }, "eval_RANGE_END": { "f1": 0.6857142857142857, "number": 82, "precision": 0.5625, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.9566759564225994, "number": 5755, "precision": 0.9306605323693723, "recall": 0.9841876629018245 }, "eval_loss": 0.4935953915119171, "eval_overall_accuracy": 0.8338845661192574, "eval_overall_f1": 0.8569683810557741, "eval_overall_precision": 0.83311345646438, "eval_overall_recall": 0.8822296730930428, "eval_runtime": 7.6552, "eval_samples_per_second": 1111.139, "eval_steps_per_second": 34.748, "step": 49000 }, { "epoch": 9.8, "learning_rate": 9.998020194020988e-07, "loss": 0.3986, "step": 49500 }, { "epoch": 9.9, "learning_rate": 5.048505246485845e-07, "loss": 0.4109, "step": 50000 }, { "epoch": 9.9, "eval_COMMENT": { "f1": 0.7082043875937006, "number": 6824, "precision": 0.6672281974860698, "recall": 0.7545427901524033 }, "eval_NAME": { "f1": 0.8094707520891365, "number": 8803, "precision": 0.794249480704056, "recall": 0.8252868340338521 }, "eval_QTY": { "f1": 0.986230876216968, "number": 7168, "precision": 0.983222407099279, "recall": 0.9892578125 }, "eval_RANGE_END": { "f1": 0.6889952153110047, "number": 82, "precision": 0.5669291338582677, "recall": 0.8780487804878049 }, "eval_UNIT": { "f1": 0.956595169734842, "number": 5755, "precision": 0.9305076392311483, "recall": 0.9841876629018245 }, "eval_loss": 0.49291422963142395, "eval_overall_accuracy": 0.8339228019194371, "eval_overall_f1": 0.8567597841213808, "eval_overall_precision": 0.8333113238692638, "eval_overall_recall": 0.8815660799105896, "eval_runtime": 7.4834, "eval_samples_per_second": 1136.652, "eval_steps_per_second": 35.545, "step": 50000 }, { "epoch": 10.0, "learning_rate": 9.899029895070283e-09, "loss": 0.4034, "step": 50500 }, { "epoch": 10.0, "step": 50510, "total_flos": 1557854363345160.0, "train_loss": 0.46329507851359913, "train_runtime": 1951.481, "train_samples_per_second": 828.104, "train_steps_per_second": 25.883 } ], "logging_steps": 500, "max_steps": 50510, "num_train_epochs": 10, "save_steps": 1000, "total_flos": 1557854363345160.0, "trial_name": null, "trial_params": null }