diff --git "a/checkpoint-2673/trainer_state.json" "b/checkpoint-2673/trainer_state.json" --- "a/checkpoint-2673/trainer_state.json" +++ "b/checkpoint-2673/trainer_state.json" @@ -11,2114 +11,2114 @@ { "epoch": 1.0, "eval_B-Claim": { - "f1-score": 0.17153996101364524, - "precision": 0.25287356321839083, - "recall": 0.12979351032448377, - "support": 339.0 + "f1-score": 0.37160751565762, + "precision": 0.42788461538461536, + "recall": 0.3284132841328413, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.08383233532934131, - "precision": 1.0, - "recall": 0.04375, - "support": 160.0 + "f1-score": 0.3578947368421052, + "precision": 0.6666666666666666, + "recall": 0.2446043165467626, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8469620831429876, - "precision": 0.7427884615384616, - "recall": 0.9851222104144527, - "support": 941.0 + "f1-score": 0.8640915593705293, + "precision": 0.7895424836601307, + "recall": 0.9541864139020537, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.5420081967213116, - "precision": 0.6803858520900321, - "recall": 0.4504044274159217, - "support": 4698.0 + "f1-score": 0.5003402749421533, + "precision": 0.5493126120741183, + "recall": 0.4593851537115721, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8343351767251743, - "precision": 0.8141717503519474, - "recall": 0.8555226824457594, - "support": 2028.0 + "f1-score": 0.7718093699515347, + "precision": 0.6502211636611093, + "recall": 0.9493293591654247, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9004323458767015, - "precision": 0.8590809093131264, - "recall": 0.9459659511472983, - "support": 14861.0 + "f1-score": 0.875016720916752, + "precision": 0.8846812731043188, + "recall": 0.865561044460127, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9988389078242631, - "precision": 0.9991008541885209, - "recall": 0.9985770987793006, - "support": 13353.0 + "f1-score": 0.9992483530087988, + "precision": 0.9995577178239717, + "recall": 0.998939179632249, + "support": 11312.0 }, - "eval_accuracy": 0.8856789444749863, - "eval_loss": 0.28347474336624146, + "eval_accuracy": 0.8614038040733883, + "eval_loss": 0.31713685393333435, "eval_macro avg": { - "f1-score": 0.6254212866619178, - "precision": 0.76405734152864, - "recall": 0.629876554361031, - "support": 36380.0 - }, - "eval_runtime": 5.0448, - "eval_samples_per_second": 16.056, - "eval_steps_per_second": 2.18, + "f1-score": 0.6771440758127848, + "precision": 0.7096952189107044, + "recall": 0.685774107364433, + "support": 29705.0 + }, + "eval_runtime": 4.8338, + "eval_samples_per_second": 16.55, + "eval_steps_per_second": 2.069, "eval_weighted avg": { - "f1-score": 0.8748148382495958, - "precision": 0.8768575102351055, - "recall": 0.8856789444749863, - "support": 36380.0 + "f1-score": 0.8576207231627551, + "precision": 0.8601529227027923, + "recall": 0.8614038040733883, + "support": 29705.0 }, "step": 81 }, { "epoch": 2.0, "eval_B-Claim": { - "f1-score": 0.5137931034482759, - "precision": 0.6182572614107884, - "recall": 0.43952802359882004, - "support": 339.0 + "f1-score": 0.4708624708624709, + "precision": 0.6392405063291139, + "recall": 0.3726937269372694, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.7622641509433963, - "precision": 0.9619047619047619, - "recall": 0.63125, - "support": 160.0 + "f1-score": 0.796875, + "precision": 0.8717948717948718, + "recall": 0.7338129496402878, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8853910477127397, - "precision": 0.8241758241758241, - "recall": 0.9564293304994687, - "support": 941.0 + "f1-score": 0.8736616702355461, + "precision": 0.796875, + "recall": 0.966824644549763, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.5413588470388762, - "precision": 0.6279853891542568, - "recall": 0.47573435504469985, - "support": 4698.0 + "f1-score": 0.5100589925881107, + "precision": 0.6459770114942529, + "recall": 0.4213946513371657, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.7949260042283298, - "precision": 0.856492027334852, - "recall": 0.7416173570019724, - "support": 2028.0 + "f1-score": 0.8401387776888176, + "precision": 0.9077277970011534, + "recall": 0.7819175360158966, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9007064868336545, - "precision": 0.8614779777627618, - "recall": 0.9436780835744566, - "support": 14861.0 + "f1-score": 0.8912891699864469, + "precision": 0.8338584492430646, + "recall": 0.9572159491884262, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9988389947942025, - "precision": 0.9990260713215463, - "recall": 0.9986519883172321, - "support": 13353.0 + "f1-score": 0.9996904982977407, + "precision": 1.0, + "recall": 0.9993811881188119, + "support": 11312.0 }, - "eval_accuracy": 0.8864211105002748, - "eval_loss": 0.2809496521949768, + "eval_accuracy": 0.8830499915839084, + "eval_loss": 0.2966194748878479, "eval_macro avg": { - "f1-score": 0.7710398049999251, - "precision": 0.8213313304378275, - "recall": 0.7409841625766643, - "support": 36380.0 - }, - "eval_runtime": 5.0579, - "eval_samples_per_second": 16.014, - "eval_steps_per_second": 2.175, + "f1-score": 0.7689395113798762, + "precision": 0.8136390908374939, + "recall": 0.7476058065410885, + "support": 29705.0 + }, + "eval_runtime": 4.8625, + "eval_samples_per_second": 16.452, + "eval_steps_per_second": 2.057, "eval_weighted avg": { - "f1-score": 0.8798131143596267, - "precision": 0.8787439244541853, - "recall": 0.8864211105002748, - "support": 36380.0 + "f1-score": 0.8731020208182413, + "precision": 0.874440834821272, + "recall": 0.8830499915839084, + "support": 29705.0 }, "step": 162 }, { "epoch": 3.0, "eval_B-Claim": { - "f1-score": 0.6401137980085349, - "precision": 0.6181318681318682, - "recall": 0.6637168141592921, - "support": 339.0 + "f1-score": 0.6085192697768763, + "precision": 0.6756756756756757, + "recall": 0.5535055350553506, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.7918781725888324, - "precision": 0.6666666666666666, - "recall": 0.975, - "support": 160.0 + "f1-score": 0.8571428571428571, + "precision": 0.851063829787234, + "recall": 0.8633093525179856, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8866442199775534, - "precision": 0.93935790725327, - "recall": 0.8395324123273114, - "support": 941.0 + "f1-score": 0.8834729626808834, + "precision": 0.8529411764705882, + "recall": 0.9162717219589257, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6254390633315594, - "precision": 0.6255056418990845, - "recall": 0.6253724989357173, - "support": 4698.0 + "f1-score": 0.5764474423833614, + "precision": 0.6584269662921348, + "recall": 0.5126218445388653, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.76480605487228, - "precision": 0.6205096714768191, - "recall": 0.9965483234714004, - "support": 2028.0 + "f1-score": 0.8581151832460733, + "precision": 0.9070282235749861, + "recall": 0.8142076502732241, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9049205625504156, - "precision": 0.9449897450922942, - "recall": 0.868111163447951, - "support": 14861.0 + "f1-score": 0.8959744247675935, + "precision": 0.8563158317922328, + "recall": 0.939484827099506, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9988758992805756, - "precision": 0.9995500562429697, - "recall": 0.9982026510896428, - "support": 13353.0 + "f1-score": 0.9996020340481981, + "precision": 1.0, + "recall": 0.9992043847241867, + "support": 11312.0 }, - "eval_accuracy": 0.8894997251236944, - "eval_loss": 0.318760484457016, + "eval_accuracy": 0.8918700555462044, + "eval_loss": 0.2552729547023773, "eval_macro avg": { - "f1-score": 0.8018111100871074, - "precision": 0.7735302223947104, - "recall": 0.852354837633045, - "support": 36380.0 - }, - "eval_runtime": 5.0862, - "eval_samples_per_second": 15.925, - "eval_steps_per_second": 2.163, + "f1-score": 0.811324882006549, + "precision": 0.8287788147989789, + "recall": 0.7998007594525776, + "support": 29705.0 + }, + "eval_runtime": 4.8422, + "eval_samples_per_second": 16.522, + "eval_steps_per_second": 2.065, "eval_weighted avg": { - "f1-score": 0.8920666591013163, - "precision": 0.9012548868310761, - "recall": 0.8894997251236944, - "support": 36380.0 + "f1-score": 0.8867633844066056, + "precision": 0.886070631898416, + "recall": 0.8918700555462044, + "support": 29705.0 }, "step": 243 }, { "epoch": 4.0, "eval_B-Claim": { - "f1-score": 0.7034883720930232, - "precision": 0.6934097421203438, - "recall": 0.7138643067846607, - "support": 339.0 + "f1-score": 0.6722408026755852, + "precision": 0.6146788990825688, + "recall": 0.7416974169741697, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.8958990536277602, - "precision": 0.9044585987261147, - "recall": 0.8875, - "support": 160.0 + "f1-score": 0.8664259927797834, + "precision": 0.8695652173913043, + "recall": 0.8633093525179856, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9051172707889126, - "precision": 0.9080213903743316, - "recall": 0.9022316684378321, - "support": 941.0 + "f1-score": 0.8687035507844755, + "precision": 0.9100346020761245, + "recall": 0.8309636650868878, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6932699265998139, - "precision": 0.6739698492462312, - "recall": 0.7137079608343976, - "support": 4698.0 + "f1-score": 0.6483151400094921, + "precision": 0.6171222046532641, + "recall": 0.6828292926768308, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8904009720534629, - "precision": 0.877815045519885, - "recall": 0.903353057199211, - "support": 2028.0 + "f1-score": 0.8696993060909791, + "precision": 0.9009584664536742, + "recall": 0.8405365126676602, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9127480293558032, - "precision": 0.9218310342461052, - "recall": 0.9038422717179194, - "support": 14861.0 + "f1-score": 0.8906159274643798, + "precision": 0.9020175517958925, + "recall": 0.879498941425547, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9981619715668254, - "precision": 0.9999248459341651, - "recall": 0.9964053021792856, - "support": 13353.0 + "f1-score": 0.999557835160948, + "precision": 0.9999115357395613, + "recall": 0.9992043847241867, + "support": 11312.0 }, - "eval_accuracy": 0.9113523914238593, - "eval_loss": 0.2517726719379425, + "eval_accuracy": 0.8935869382258879, + "eval_loss": 0.29465603828430176, "eval_macro avg": { - "f1-score": 0.8570122280122289, - "precision": 0.8542043580238825, - "recall": 0.8601292238790438, - "support": 36380.0 - }, - "eval_runtime": 5.062, - "eval_samples_per_second": 16.002, - "eval_steps_per_second": 2.173, + "f1-score": 0.8307940792808061, + "precision": 0.8306126395989127, + "recall": 0.8340056522961811, + "support": 29705.0 + }, + "eval_runtime": 4.8494, + "eval_samples_per_second": 16.497, + "eval_steps_per_second": 2.062, "eval_weighted avg": { - "f1-score": 0.9122886837716081, - "precision": 0.9134709768686261, - "recall": 0.9113523914238593, - "support": 36380.0 + "f1-score": 0.8954766464091573, + "precision": 0.8982496227307208, + "recall": 0.8935869382258879, + "support": 29705.0 }, "step": 324 }, { "epoch": 5.0, "eval_B-Claim": { - "f1-score": 0.6590538336052201, - "precision": 0.7372262773722628, - "recall": 0.5958702064896755, - "support": 339.0 + "f1-score": 0.6438095238095239, + "precision": 0.6653543307086615, + "recall": 0.6236162361623616, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9078947368421054, - "precision": 0.9583333333333334, - "recall": 0.8625, - "support": 160.0 + "f1-score": 0.8785714285714286, + "precision": 0.8723404255319149, + "recall": 0.8848920863309353, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9066802651708312, - "precision": 0.8715686274509804, - "recall": 0.944739638682253, - "support": 941.0 + "f1-score": 0.8783151326053042, + "precision": 0.8674884437596302, + "recall": 0.8894154818325435, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6513589503280225, - "precision": 0.7243355914538823, - "recall": 0.5917411664538101, - "support": 4698.0 + "f1-score": 0.6283729628640129, + "precision": 0.6748923959827834, + "recall": 0.5878530367408148, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8981975120588981, - "precision": 0.9256933542647828, - "recall": 0.8722879684418146, - "support": 2028.0 + "f1-score": 0.885450461692039, + "precision": 0.8896690070210632, + "recall": 0.8812717337307501, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9158848058378942, - "precision": 0.8877170824123777, - "recall": 0.9458986609245676, - "support": 14861.0 + "f1-score": 0.8989315871101154, + "precision": 0.878494442573257, + "recall": 0.9203422724064926, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9994384336041331, - "precision": 0.9992513849378649, - "recall": 0.9996255523103422, - "support": 13353.0 + "f1-score": 0.9997347245556637, + "precision": 1.0, + "recall": 0.9994695898161244, + "support": 11312.0 }, - "eval_accuracy": 0.9121220450797142, - "eval_loss": 0.2739432156085968, + "eval_accuracy": 0.8995118666891095, + "eval_loss": 0.3176642060279846, "eval_macro avg": { - "f1-score": 0.8483583624924435, - "precision": 0.8720179501750691, - "recall": 0.8303804561860663, - "support": 36380.0 - }, - "eval_runtime": 5.0932, - "eval_samples_per_second": 15.904, - "eval_steps_per_second": 2.16, + "f1-score": 0.830455117315441, + "precision": 0.8354627207967587, + "recall": 0.8266943481457174, + "support": 29705.0 + }, + "eval_runtime": 4.8682, + "eval_samples_per_second": 16.433, + "eval_steps_per_second": 2.054, "eval_weighted avg": { - "f1-score": 0.9087400479269403, - "precision": 0.9081638580455985, - "recall": 0.9121220450797142, - "support": 36380.0 + "f1-score": 0.8971010593476484, + "precision": 0.8958911872123141, + "recall": 0.8995118666891095, + "support": 29705.0 }, "step": 405 }, { "epoch": 6.0, "eval_B-Claim": { - "f1-score": 0.6675461741424802, - "precision": 0.60381861575179, - "recall": 0.7463126843657817, - "support": 339.0 + "f1-score": 0.6150712830957231, + "precision": 0.6863636363636364, + "recall": 0.5571955719557196, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.887608069164265, - "precision": 0.8235294117647058, - "recall": 0.9625, - "support": 160.0 + "f1-score": 0.8686131386861314, + "precision": 0.8814814814814815, + "recall": 0.8561151079136691, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8737316798196167, - "precision": 0.9303721488595438, - "recall": 0.8235919234856536, - "support": 941.0 + "f1-score": 0.8804841149773072, + "precision": 0.8447024673439768, + "recall": 0.919431279620853, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6529236868186323, - "precision": 0.6109050445103857, - "recall": 0.7011494252873564, - "support": 4698.0 + "f1-score": 0.5758052970651396, + "precision": 0.6739276139410187, + "recall": 0.5026243439140214, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8689320388349515, - "precision": 0.7863418530351438, - "recall": 0.9709072978303748, - "support": 2028.0 + "f1-score": 0.8627552339105711, + "precision": 0.8992456896551724, + "recall": 0.829110779930452, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.8937734066010992, - "precision": 0.9312910284463894, - "recall": 0.8591615638247763, - "support": 14861.0 + "f1-score": 0.894107779408681, + "precision": 0.8521864257734432, + "recall": 0.9403669724770642, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9988009592326139, - "precision": 0.9994750656167979, - "recall": 0.9981277615517112, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.8944749862561847, - "eval_loss": 0.37784042954444885, + "eval_accuracy": 0.8922403635751557, + "eval_loss": 0.41071853041648865, "eval_macro avg": { - "f1-score": 0.8347594306590942, - "precision": 0.8122475954263937, - "recall": 0.8659643794779507, - "support": 36380.0 - }, - "eval_runtime": 5.0734, - "eval_samples_per_second": 15.966, - "eval_steps_per_second": 2.168, + "f1-score": 0.8138338353062219, + "precision": 0.8339867592226755, + "recall": 0.8006920079731114, + "support": 29705.0 + }, + "eval_runtime": 4.8571, + "eval_samples_per_second": 16.471, + "eval_steps_per_second": 2.059, "eval_weighted avg": { - "f1-score": 0.8971820344797716, - "precision": 0.9033144340485707, - "recall": 0.8944749862561847, - "support": 36380.0 + "f1-score": 0.8864802913843919, + "precision": 0.8861194550557427, + "recall": 0.8922403635751557, + "support": 29705.0 }, "step": 486 }, { "epoch": 6.17, - "grad_norm": 21.778270721435547, + "grad_norm": 5.239528656005859, "learning_rate": 1.7530864197530865e-05, - "loss": 0.2481, + "loss": 0.2363, "step": 500 }, { "epoch": 7.0, "eval_B-Claim": { - "f1-score": 0.6489859594383774, - "precision": 0.6887417218543046, - "recall": 0.6135693215339233, - "support": 339.0 + "f1-score": 0.6954954954954955, + "precision": 0.6795774647887324, + "recall": 0.7121771217712177, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.8852459016393444, - "precision": 0.9310344827586207, - "recall": 0.84375, - "support": 160.0 + "f1-score": 0.8920863309352518, + "precision": 0.8920863309352518, + "recall": 0.8920863309352518, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9011898603207449, - "precision": 0.8780241935483871, - "recall": 0.9256110520722636, - "support": 941.0 + "f1-score": 0.8876494023904382, + "precision": 0.8954983922829582, + "recall": 0.8799368088467614, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.626564114338193, - "precision": 0.6800398704211313, - "recall": 0.5808854831843338, - "support": 4698.0 + "f1-score": 0.6831545741324923, + "precision": 0.6898572884811417, + "recall": 0.6765808547863035, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8813559322033898, - "precision": 0.9049350649350649, - "recall": 0.8589743589743589, - "support": 2028.0 + "f1-score": 0.8984707946853848, + "precision": 0.9068825910931174, + "recall": 0.8902136115250869, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9073595340161655, - "precision": 0.8831698305516626, - "recall": 0.9329116479375547, - "support": 14861.0 + "f1-score": 0.9068164859763859, + "precision": 0.9024198479951079, + "recall": 0.9112561750176429, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9981994148098132, - "precision": 1.0, - "recall": 0.9964053021792856, - "support": 13353.0 + "f1-score": 0.9999558011049724, + "precision": 0.9999116061168567, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9030786146234194, - "eval_loss": 0.4109443426132202, + "eval_accuracy": 0.9094428547382596, + "eval_loss": 0.3520519435405731, "eval_macro avg": { - "f1-score": 0.8355572452522898, - "precision": 0.8522778805813102, - "recall": 0.8217295951259599, - "support": 36380.0 - }, - "eval_runtime": 5.0454, - "eval_samples_per_second": 16.054, - "eval_steps_per_second": 2.18, + "f1-score": 0.8519469835314888, + "precision": 0.8523190745275951, + "recall": 0.8517501289831806, + "support": 29705.0 + }, + "eval_runtime": 4.8905, + "eval_samples_per_second": 16.358, + "eval_steps_per_second": 2.045, "eval_weighted avg": { - "f1-score": 0.9003265559019435, - "precision": 0.8992988510674564, - "recall": 0.9030786146234194, - "support": 36380.0 + "f1-score": 0.9091888981291354, + "precision": 0.908989097041669, + "recall": 0.9094428547382596, + "support": 29705.0 }, "step": 567 }, { "epoch": 8.0, "eval_B-Claim": { - "f1-score": 0.7029972752043596, - "precision": 0.6531645569620254, - "recall": 0.7610619469026548, - "support": 339.0 + "f1-score": 0.6666666666666667, + "precision": 0.658273381294964, + "recall": 0.6752767527675276, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9134328358208955, - "precision": 0.8742857142857143, - "recall": 0.95625, - "support": 160.0 + "f1-score": 0.8819444444444444, + "precision": 0.8523489932885906, + "recall": 0.9136690647482014, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8914728682170544, - "precision": 0.930635838150289, - "recall": 0.8554729011689692, - "support": 941.0 + "f1-score": 0.8759007205764612, + "precision": 0.887987012987013, + "recall": 0.8641390205371248, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6790487421383647, - "precision": 0.6307046367287331, - "recall": 0.7354193273733504, - "support": 4698.0 + "f1-score": 0.642274412855377, + "precision": 0.6353631694790902, + "recall": 0.6493376655836041, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8939393939393939, - "precision": 0.836340206185567, - "recall": 0.9600591715976331, - "support": 2028.0 + "f1-score": 0.87683284457478, + "precision": 0.862914862914863, + "recall": 0.8912071535022354, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.8993180625983564, - "precision": 0.9362166885102665, - "recall": 0.8652176838705337, - "support": 14861.0 + "f1-score": 0.8911585772014743, + "precision": 0.8972547616918537, + "recall": 0.8851446718419196, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9979819119515658, - "precision": 0.9960462513987318, - "recall": 0.9999251104620684, - "support": 13353.0 + "f1-score": 0.9998673915926268, + "precision": 0.9999115904871364, + "recall": 0.9998231966053748, + "support": 11312.0 }, - "eval_accuracy": 0.902363936228697, - "eval_loss": 0.4763801097869873, + "eval_accuracy": 0.8952364921730348, + "eval_loss": 0.4631531238555908, "eval_macro avg": { - "f1-score": 0.8540272985528559, - "precision": 0.8367705560316182, - "recall": 0.8762008773393156, - "support": 36380.0 - }, - "eval_runtime": 5.0791, - "eval_samples_per_second": 15.948, - "eval_steps_per_second": 2.166, + "f1-score": 0.8335207225588329, + "precision": 0.827721967449073, + "recall": 0.8397996465122839, + "support": 29705.0 + }, + "eval_runtime": 4.8517, + "eval_samples_per_second": 16.489, + "eval_steps_per_second": 2.061, "eval_weighted avg": { - "f1-score": 0.9048169208096876, - "precision": 0.9101018951943243, - "recall": 0.902363936228697, - "support": 36380.0 + "f1-score": 0.8956465277692122, + "precision": 0.8961582200263131, + "recall": 0.8952364921730348, + "support": 29705.0 }, "step": 648 }, { "epoch": 9.0, "eval_B-Claim": { - "f1-score": 0.6884779516358464, - "precision": 0.6648351648351648, - "recall": 0.7138643067846607, - "support": 339.0 + "f1-score": 0.6604127579737337, + "precision": 0.6717557251908397, + "recall": 0.6494464944649446, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.8860759493670887, - "precision": 0.8974358974358975, - "recall": 0.875, - "support": 160.0 + "f1-score": 0.8661417322834645, + "precision": 0.9565217391304348, + "recall": 0.7913669064748201, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8984416980118215, - "precision": 0.908695652173913, - "recall": 0.8884165781083954, - "support": 941.0 + "f1-score": 0.8852963818321786, + "precision": 0.8633633633633634, + "recall": 0.9083728278041074, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6685829873309911, - "precision": 0.6687965921192758, - "recall": 0.6683695189442316, - "support": 4698.0 + "f1-score": 0.6148568412719356, + "precision": 0.6512017887087759, + "recall": 0.5823544113971507, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8819011041766683, - "precision": 0.8592142188961647, - "recall": 0.9058185404339251, - "support": 2028.0 + "f1-score": 0.8500414250207126, + "precision": 0.957089552238806, + "recall": 0.7645305514157973, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9072658065820057, - "precision": 0.9083367057871308, - "recall": 0.9061974295134917, - "support": 14861.0 + "f1-score": 0.8985408601693112, + "precision": 0.8677183468901487, + "recall": 0.9316337332392378, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9971465044679734, - "precision": 0.9998494089300505, - "recall": 0.9944581741930653, - "support": 13353.0 + "f1-score": 0.9995136401821638, + "precision": 0.9998230871295887, + "recall": 0.9992043847241867, + "support": 11312.0 }, - "eval_accuracy": 0.9054700384826828, - "eval_loss": 0.43875375390052795, + "eval_accuracy": 0.8952701565393032, + "eval_loss": 0.5578464865684509, "eval_macro avg": { - "f1-score": 0.8468417145103421, - "precision": 0.843880520025371, - "recall": 0.8503035068539672, - "support": 36380.0 - }, - "eval_runtime": 5.0863, - "eval_samples_per_second": 15.925, - "eval_steps_per_second": 2.163, + "f1-score": 0.8249719483905, + "precision": 0.8524962289502797, + "recall": 0.8038441870743206, + "support": 29705.0 + }, + "eval_runtime": 4.8325, + "eval_samples_per_second": 16.555, + "eval_steps_per_second": 2.069, "eval_weighted avg": { - "f1-score": 0.9056589487317133, - "precision": 0.9059462683069767, - "recall": 0.9054700384826828, - "support": 36380.0 + "f1-score": 0.8928897918536849, + "precision": 0.8934537922744324, + "recall": 0.8952701565393032, + "support": 29705.0 }, "step": 729 }, { "epoch": 10.0, "eval_B-Claim": { - "f1-score": 0.6910569105691057, - "precision": 0.6390977443609023, - "recall": 0.7522123893805309, - "support": 339.0 + "f1-score": 0.6775956284153005, + "precision": 0.6690647482014388, + "recall": 0.6863468634686347, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9046153846153846, - "precision": 0.8909090909090909, - "recall": 0.91875, - "support": 160.0 + "f1-score": 0.8832116788321168, + "precision": 0.8962962962962963, + "recall": 0.8705035971223022, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8888888888888888, - "precision": 0.9213226909920182, - "recall": 0.8586609989373007, - "support": 941.0 + "f1-score": 0.8820269200316706, + "precision": 0.8841269841269841, + "recall": 0.8799368088467614, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6645316253002403, - "precision": 0.6271250472232717, - "recall": 0.7066836951894423, - "support": 4698.0 + "f1-score": 0.6348430262480701, + "precision": 0.6542031291434632, + "recall": 0.6165958510372407, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.886980737990253, - "precision": 0.8377904427882508, - "recall": 0.9423076923076923, - "support": 2028.0 + "f1-score": 0.8864321608040201, + "precision": 0.896797153024911, + "recall": 0.8763040238450075, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.8983343700324833, - "precision": 0.9233501456276195, - "recall": 0.8746383150528229, - "support": 14861.0 + "f1-score": 0.8934983440822729, + "precision": 0.8828797795384086, + "recall": 0.9043754410726887, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9975225225225225, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9950572904965176, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.8995601979109401, - "eval_loss": 0.5381875038146973, + "eval_accuracy": 0.897458340346743, + "eval_loss": 0.6301500201225281, "eval_macro avg": { - "f1-score": 0.8474186342741253, - "precision": 0.8342278802715933, - "recall": 0.8640443401949011, - "support": 36380.0 - }, - "eval_runtime": 5.099, - "eval_samples_per_second": 15.885, - "eval_steps_per_second": 2.157, + "f1-score": 0.8368011083447787, + "precision": 0.8404811557616432, + "recall": 0.8334375121989479, + "support": 29705.0 + }, + "eval_runtime": 4.8426, + "eval_samples_per_second": 16.52, + "eval_steps_per_second": 2.065, "eval_weighted avg": { - "f1-score": 0.9017669881051834, - "precision": 0.9056169116577473, - "recall": 0.8995601979109401, - "support": 36380.0 + "f1-score": 0.8964756773362035, + "precision": 0.895761611933671, + "recall": 0.897458340346743, + "support": 29705.0 }, "step": 810 }, { "epoch": 11.0, "eval_B-Claim": { - "f1-score": 0.7103825136612023, - "precision": 0.6615776081424937, - "recall": 0.7669616519174042, - "support": 339.0 + "f1-score": 0.6483300589390962, + "precision": 0.6932773109243697, + "recall": 0.6088560885608856, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9032258064516129, - "precision": 0.850828729281768, - "recall": 0.9625, - "support": 160.0 + "f1-score": 0.8805970149253732, + "precision": 0.9147286821705426, + "recall": 0.8489208633093526, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8955223880597015, - "precision": 0.9331797235023042, - "recall": 0.8607863974495218, - "support": 941.0 + "f1-score": 0.8848207475209764, + "precision": 0.855457227138643, + "recall": 0.9162717219589257, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.685932169375878, - "precision": 0.6488230827638573, - "recall": 0.7275436355896125, - "support": 4698.0 + "f1-score": 0.6216666666666667, + "precision": 0.6995936230071897, + "recall": 0.55936015996001, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8921899617375647, - "precision": 0.820703933747412, - "recall": 0.9773175542406312, - "support": 2028.0 + "f1-score": 0.8927083333333333, + "precision": 0.9381499726327313, + "recall": 0.851465474416294, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9036754507628294, - "precision": 0.932183990271121, - "recall": 0.8768588924029338, - "support": 14861.0 + "f1-score": 0.9015977681968045, + "precision": 0.8655250770978737, + "recall": 0.9408080451658434, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9971084156370874, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9942335055792706, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9051951621770203, - "eval_loss": 0.5869538187980652, + "eval_accuracy": 0.9019357010604275, + "eval_loss": 0.5967662334442139, "eval_macro avg": { - "f1-score": 0.8554338150979823, - "precision": 0.8353281525298509, - "recall": 0.880885948168482, - "support": 36380.0 - }, - "eval_runtime": 5.0923, - "eval_samples_per_second": 15.906, - "eval_steps_per_second": 2.16, + "f1-score": 0.8328172270831785, + "precision": 0.8523902704244785, + "recall": 0.8179546219101873, + "support": 29705.0 + }, + "eval_runtime": 4.8502, + "eval_samples_per_second": 16.494, + "eval_steps_per_second": 2.062, "eval_weighted avg": { - "f1-score": 0.9071963301332511, - "precision": 0.9114149044954519, - "recall": 0.9051951621770203, - "support": 36380.0 + "f1-score": 0.8979974574260295, + "precision": 0.8977509549706758, + "recall": 0.9019357010604275, + "support": 29705.0 }, "step": 891 }, { "epoch": 12.0, "eval_B-Claim": { - "f1-score": 0.6758409785932722, - "precision": 0.7015873015873015, - "recall": 0.6519174041297935, - "support": 339.0 + "f1-score": 0.6921898928024502, + "precision": 0.5916230366492147, + "recall": 0.8339483394833949, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.916923076923077, - "precision": 0.9030303030303031, - "recall": 0.93125, - "support": 160.0 + "f1-score": 0.8978102189781022, + "precision": 0.9111111111111111, + "recall": 0.8848920863309353, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8989473684210527, - "precision": 0.8905109489051095, - "recall": 0.9075451647183846, - "support": 941.0 + "f1-score": 0.8482758620689655, + "precision": 0.9335863377609108, + "recall": 0.7772511848341233, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6533197602081212, - "precision": 0.6970794110547912, - "recall": 0.6147296722009365, - "support": 4698.0 + "f1-score": 0.6750635055038103, + "precision": 0.5854598861758766, + "recall": 0.7970507373156711, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9200480192076831, - "precision": 0.8965839962564343, - "recall": 0.9447731755424064, - "support": 2028.0 + "f1-score": 0.8948170731707318, + "precision": 0.9157566302652106, + "recall": 0.8748137108792846, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9082653940478558, - "precision": 0.8948605759811924, - "recall": 0.922077922077922, - "support": 14861.0 + "f1-score": 0.8742200328407225, + "precision": 0.9336606874436316, + "recall": 0.821894848270995, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9988389947942025, - "precision": 0.9990260713215463, - "recall": 0.9986519883172321, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9089059923034635, - "eval_loss": 0.5580677390098572, + "eval_accuracy": 0.8894125568086181, + "eval_loss": 0.6401852965354919, "eval_macro avg": { - "f1-score": 0.8531690845993235, - "precision": 0.854668372590954, - "recall": 0.852992189569525, - "support": 36380.0 - }, - "eval_runtime": 5.0574, - "eval_samples_per_second": 16.016, - "eval_steps_per_second": 2.175, + "f1-score": 0.8403395121949689, + "precision": 0.8387425270579937, + "recall": 0.8556929867306292, + "support": 29705.0 + }, + "eval_runtime": 4.8727, + "eval_samples_per_second": 16.418, + "eval_steps_per_second": 2.052, "eval_weighted avg": { - "f1-score": 0.9068749310769137, - "precision": 0.9057713940131727, - "recall": 0.9089059923034635, - "support": 36380.0 + "f1-score": 0.8945864954779644, + "precision": 0.9075830699664568, + "recall": 0.8894125568086181, + "support": 29705.0 }, "step": 972 }, { "epoch": 12.35, - "grad_norm": 0.45772114396095276, + "grad_norm": 23.30899429321289, "learning_rate": 1.506172839506173e-05, - "loss": 0.0314, + "loss": 0.0272, "step": 1000 }, { "epoch": 13.0, "eval_B-Claim": { - "f1-score": 0.6760563380281691, - "precision": 0.72, - "recall": 0.6371681415929203, - "support": 339.0 + "f1-score": 0.706766917293233, + "precision": 0.7203065134099617, + "recall": 0.6937269372693727, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9171974522292993, - "precision": 0.935064935064935, - "recall": 0.9, - "support": 160.0 + "f1-score": 0.8958333333333333, + "precision": 0.8657718120805369, + "recall": 0.9280575539568345, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.905503634475597, - "precision": 0.8852791878172589, - "recall": 0.926673751328374, - "support": 941.0 + "f1-score": 0.8957345971563981, + "precision": 0.8957345971563981, + "recall": 0.8957345971563981, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6522588803310726, - "precision": 0.7090727318170458, - "recall": 0.6038739889314602, - "support": 4698.0 + "f1-score": 0.6850271528316524, + "precision": 0.7096169300830432, + "recall": 0.6620844788802799, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9228121927236971, - "precision": 0.9200980392156862, - "recall": 0.9255424063116371, - "support": 2028.0 + "f1-score": 0.8991678903573177, + "precision": 0.8861553304389773, + "recall": 0.912568306010929, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9102441910145594, - "precision": 0.8896312475909033, - "recall": 0.9318350043738645, - "support": 14861.0 + "f1-score": 0.9089637690660374, + "precision": 0.9007362494586401, + "recall": 0.9173429781227946, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9992880428673137, + "f1-score": 0.9999557971975424, "precision": 1.0, - "recall": 0.9985770987793006, - "support": 13353.0 + "recall": 0.9999115983026874, + "support": 11312.0 }, - "eval_accuracy": 0.9106102253985706, - "eval_loss": 0.5264647603034973, + "eval_accuracy": 0.9116310385456994, + "eval_loss": 0.6060317754745483, "eval_macro avg": { - "f1-score": 0.8547658188099584, - "precision": 0.8655923059294041, - "recall": 0.8462386273310795, - "support": 36380.0 - }, - "eval_runtime": 5.0562, - "eval_samples_per_second": 16.02, - "eval_steps_per_second": 2.176, + "f1-score": 0.855921351033645, + "precision": 0.8540459189467938, + "recall": 0.8584894928141853, + "support": 29705.0 + }, + "eval_runtime": 4.8344, + "eval_samples_per_second": 16.548, + "eval_steps_per_second": 2.068, "eval_weighted avg": { - "f1-score": 0.9080379636257123, - "precision": 0.9070294280758862, - "recall": 0.9106102253985706, - "support": 36380.0 + "f1-score": 0.9106004556040976, + "precision": 0.9098905735839876, + "recall": 0.9116310385456994, + "support": 29705.0 }, "step": 1053 }, { "epoch": 14.0, "eval_B-Claim": { - "f1-score": 0.6980609418282548, - "precision": 0.6579634464751958, - "recall": 0.7433628318584071, - "support": 339.0 + "f1-score": 0.6832740213523131, + "precision": 0.6597938144329897, + "recall": 0.7084870848708487, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9226006191950465, - "precision": 0.9141104294478528, - "recall": 0.93125, - "support": 160.0 + "f1-score": 0.9044117647058822, + "precision": 0.924812030075188, + "recall": 0.8848920863309353, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8921568627450981, - "precision": 0.9150837988826815, - "recall": 0.8703506907545164, - "support": 941.0 + "f1-score": 0.8753993610223643, + "precision": 0.8852988691437803, + "recall": 0.8657187993680885, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6821321785751109, - "precision": 0.6614677064587082, - "recall": 0.7041294167730949, - "support": 4698.0 + "f1-score": 0.6549304444170873, + "precision": 0.6453178068898593, + "recall": 0.664833791552112, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9124910093502757, - "precision": 0.8880074661689221, - "recall": 0.9383629191321499, - "support": 2028.0 + "f1-score": 0.8969913309535952, + "precision": 0.9214248297537978, + "recall": 0.8738201689021361, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9052323204796293, - "precision": 0.9166609175577786, - "recall": 0.894085189421977, - "support": 14861.0 + "f1-score": 0.8920768042374753, + "precision": 0.8927467090732397, + "recall": 0.891407904022583, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9980114808839531, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9960308544896278, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9075865860362837, - "eval_loss": 0.6585939526557922, + "eval_accuracy": 0.8988049149974752, + "eval_loss": 0.6456504464149475, "eval_macro avg": { - "f1-score": 0.8586693447224812, - "precision": 0.850470537855877, - "recall": 0.8682245574899676, - "support": 36380.0 - }, - "eval_runtime": 5.0606, - "eval_samples_per_second": 16.006, - "eval_steps_per_second": 2.174, + "f1-score": 0.8438691038126739, + "precision": 0.8470562941955507, + "recall": 0.8413085478638147, + "support": 29705.0 + }, + "eval_runtime": 4.8614, + "eval_samples_per_second": 16.456, + "eval_steps_per_second": 2.057, "eval_weighted avg": { - "f1-score": 0.9086880683615599, - "precision": 0.9102351141829325, - "recall": 0.9075865860362837, - "support": 36380.0 + "f1-score": 0.8993641205442222, + "precision": 0.900073004604436, + "recall": 0.8988049149974752, + "support": 29705.0 }, "step": 1134 }, { "epoch": 15.0, "eval_B-Claim": { - "f1-score": 0.6920731707317073, - "precision": 0.7160883280757098, - "recall": 0.6696165191740413, - "support": 339.0 + "f1-score": 0.6766917293233083, + "precision": 0.6896551724137931, + "recall": 0.6642066420664207, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.90282131661442, - "precision": 0.9056603773584906, - "recall": 0.9, - "support": 160.0 + "f1-score": 0.899628252788104, + "precision": 0.9307692307692308, + "recall": 0.8705035971223022, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9076600209863588, - "precision": 0.8963730569948186, - "recall": 0.9192348565356004, - "support": 941.0 + "f1-score": 0.8849144634525662, + "precision": 0.8713629402756509, + "recall": 0.8988941548183255, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6644877718922575, - "precision": 0.7061077844311378, - "recall": 0.6275010642826735, - "support": 4698.0 + "f1-score": 0.6276567124024752, + "precision": 0.6795805418001748, + "recall": 0.583104223944014, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9016511127063891, - "precision": 0.8758716875871687, - "recall": 0.9289940828402367, - "support": 2028.0 + "f1-score": 0.9034411915767849, + "precision": 0.9351408825093036, + "recall": 0.8738201689021361, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9134082844161443, - "precision": 0.9005362280931206, - "recall": 0.9266536572236054, - "support": 14861.0 + "f1-score": 0.8967347113450592, + "precision": 0.8708336796608761, + "recall": 0.9242237120677488, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9987253505286046, - "precision": 0.9999249305607687, - "recall": 0.9975286452482588, - "support": 13353.0 + "f1-score": 0.9998232278592893, + "precision": 0.9996465182043125, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9114623419461243, - "eval_loss": 0.6387954950332642, + "eval_accuracy": 0.9005554620434271, + "eval_loss": 0.7282584309577942, "eval_macro avg": { - "f1-score": 0.854403861125126, - "precision": 0.8572231990144593, - "recall": 0.8527898321863451, - "support": 36380.0 - }, - "eval_runtime": 5.0835, - "eval_samples_per_second": 15.934, - "eval_steps_per_second": 2.164, + "f1-score": 0.8412700412496553, + "precision": 0.8538555665190488, + "recall": 0.8306789284172782, + "support": 29705.0 + }, + "eval_runtime": 4.8599, + "eval_samples_per_second": 16.461, + "eval_steps_per_second": 2.058, "eval_weighted avg": { - "f1-score": 0.9096653222792656, - "precision": 0.9087293921765042, - "recall": 0.9114623419461243, - "support": 36380.0 + "f1-score": 0.8979582328721724, + "precision": 0.8971236682980019, + "recall": 0.9005554620434271, + "support": 29705.0 }, "step": 1215 }, { "epoch": 16.0, "eval_B-Claim": { - "f1-score": 0.6602254428341385, - "precision": 0.7269503546099291, - "recall": 0.6047197640117994, - "support": 339.0 + "f1-score": 0.7162629757785466, + "precision": 0.6742671009771987, + "recall": 0.7638376383763837, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.8903225806451613, - "precision": 0.92, - "recall": 0.8625, - "support": 160.0 + "f1-score": 0.909090909090909, + "precision": 0.9191176470588235, + "recall": 0.8992805755395683, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9055441478439425, - "precision": 0.8758689175769613, - "recall": 0.9373007438894793, - "support": 941.0 + "f1-score": 0.884022708840227, + "precision": 0.9083333333333333, + "recall": 0.8609794628751974, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6361776387050188, - "precision": 0.7319302132373304, - "recall": 0.5625798212005109, - "support": 4698.0 + "f1-score": 0.6889387321449651, + "precision": 0.6527964205816554, + "recall": 0.7293176705823544, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9097614949594295, - "precision": 0.9073075036782736, - "recall": 0.9122287968441815, - "support": 2028.0 + "f1-score": 0.9093224656138562, + "precision": 0.9330893883951908, + "recall": 0.886736214605067, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9127908863717504, - "precision": 0.8816300940438871, - "recall": 0.9462351120382209, - "support": 14861.0 + "f1-score": 0.8969196969017621, + "precision": 0.9120087535333272, + "recall": 0.8823218066337333, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9993256911665542, - "precision": 0.9997751293006522, - "recall": 0.9988766569310268, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9103353490929081, - "eval_loss": 0.6975058317184448, + "eval_accuracy": 0.9053694664197947, + "eval_loss": 0.6488288640975952, "eval_macro avg": { - "f1-score": 0.844878268932285, - "precision": 0.8633517446352906, - "recall": 0.8320629849878884, - "support": 36380.0 - }, - "eval_runtime": 5.0731, - "eval_samples_per_second": 15.967, - "eval_steps_per_second": 2.168, + "f1-score": 0.8577939269100382, + "precision": 0.8570875205542184, + "recall": 0.8603533383731864, + "support": 29705.0 + }, + "eval_runtime": 4.862, + "eval_samples_per_second": 16.454, + "eval_steps_per_second": 2.057, "eval_weighted avg": { - "f1-score": 0.9060232340867683, - "precision": 0.9056723119856793, - "recall": 0.9103353490929081, - "support": 36380.0 + "f1-score": 0.9071351271559185, + "precision": 0.9098177877443702, + "recall": 0.9053694664197947, + "support": 29705.0 }, "step": 1296 }, { "epoch": 17.0, "eval_B-Claim": { - "f1-score": 0.6964285714285714, - "precision": 0.7027027027027027, - "recall": 0.6902654867256637, - "support": 339.0 + "f1-score": 0.7306397306397308, + "precision": 0.6718266253869969, + "recall": 0.8007380073800738, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.903954802259887, - "precision": 0.8247422680412371, - "recall": 1.0, - "support": 160.0 + "f1-score": 0.924187725631769, + "precision": 0.927536231884058, + "recall": 0.920863309352518, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9045822102425876, - "precision": 0.9179431072210066, - "recall": 0.8916046758767269, - "support": 941.0 + "f1-score": 0.8808545603944125, + "precision": 0.9178082191780822, + "recall": 0.8467614533965245, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6845814977973568, - "precision": 0.7092651757188498, - "recall": 0.6615581098339719, - "support": 4698.0 + "f1-score": 0.7016976556184318, + "precision": 0.652211249463289, + "recall": 0.7593101724568858, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.8985378821444396, - "precision": 0.8157683024939663, - "recall": 1.0, - "support": 2028.0 + "f1-score": 0.9132860040567952, + "precision": 0.9326773692387365, + "recall": 0.8946845504222554, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9151271572832584, - "precision": 0.9186330349877949, - "recall": 0.9116479375546733, - "support": 14861.0 + "f1-score": 0.8968950846383634, + "precision": 0.9209890314184793, + "recall": 0.874029640084686, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9988753936122357, - "precision": 1.0, - "recall": 0.9977533138620535, - "support": 13353.0 + "f1-score": 0.9999558011049724, + "precision": 0.9999116061168567, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9136888400219901, - "eval_loss": 0.692401111125946, + "eval_accuracy": 0.9069180272681366, + "eval_loss": 0.616534411907196, "eval_macro avg": { - "f1-score": 0.8574410735383337, - "precision": 0.8412935130236511, - "recall": 0.878975646264727, - "support": 36380.0 - }, - "eval_runtime": 5.0854, - "eval_samples_per_second": 15.928, - "eval_steps_per_second": 2.163, + "f1-score": 0.8639309374406394, + "precision": 0.8604229046694998, + "recall": 0.8709124475847062, + "support": 29705.0 + }, + "eval_runtime": 4.847, + "eval_samples_per_second": 16.505, + "eval_steps_per_second": 2.063, "eval_weighted avg": { - "f1-score": 0.9128097974394099, - "precision": 0.9132839716585025, - "recall": 0.9136888400219901, - "support": 36380.0 + "f1-score": 0.9092303028261727, + "precision": 0.913323469327906, + "recall": 0.9069180272681366, + "support": 29705.0 }, "step": 1377 }, { "epoch": 18.0, "eval_B-Claim": { - "f1-score": 0.6927899686520376, - "precision": 0.7391304347826086, - "recall": 0.6519174041297935, - "support": 339.0 + "f1-score": 0.7090909090909091, + "precision": 0.6989247311827957, + "recall": 0.7195571955719557, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9192546583850932, - "precision": 0.9135802469135802, - "recall": 0.925, - "support": 160.0 + "f1-score": 0.9257950530035336, + "precision": 0.9097222222222222, + "recall": 0.9424460431654677, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9089016137428423, - "precision": 0.8908163265306123, - "recall": 0.9277364505844846, - "support": 941.0 + "f1-score": 0.885873902633679, + "precision": 0.8951612903225806, + "recall": 0.8767772511848341, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6788553259141495, - "precision": 0.7276046738072055, - "recall": 0.6362281822051937, - "support": 4698.0 + "f1-score": 0.6834259727134917, + "precision": 0.6909323116219668, + "recall": 0.6760809797550612, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9119420989143546, - "precision": 0.8927727916863486, - "recall": 0.9319526627218935, - "support": 2028.0 + "f1-score": 0.9266351653817458, + "precision": 0.9277888446215139, + "recall": 0.9254843517138599, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9172837873578418, - "precision": 0.9016052511860662, - "recall": 0.9335172599421304, - "support": 14861.0 + "f1-score": 0.9006238467621475, + "precision": 0.8970768422895151, + "recall": 0.9041990119971771, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9990254872563717, - "precision": 1.0, - "recall": 0.9980528720137797, - "support": 13353.0 + "f1-score": 0.9999558011049724, + "precision": 0.9999116061168567, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.915915338097856, - "eval_loss": 0.6629185676574707, + "eval_accuracy": 0.9093081972731863, + "eval_loss": 0.638508677482605, "eval_macro avg": { - "f1-score": 0.861150420031813, - "precision": 0.866501389272346, - "recall": 0.8577721187996108, - "support": 36380.0 - }, - "eval_runtime": 5.0772, - "eval_samples_per_second": 15.954, - "eval_steps_per_second": 2.167, + "f1-score": 0.8616286643843541, + "precision": 0.8599311211967786, + "recall": 0.8635064047697651, + "support": 29705.0 + }, + "eval_runtime": 4.9839, + "eval_samples_per_second": 16.052, + "eval_steps_per_second": 2.006, "eval_weighted avg": { - "f1-score": 0.9138987233713269, - "precision": 0.9130176425817779, - "recall": 0.915915338097856, - "support": 36380.0 + "f1-score": 0.9090147511474005, + "precision": 0.9087634870703264, + "recall": 0.9093081972731863, + "support": 29705.0 }, "step": 1458 }, { "epoch": 18.52, - "grad_norm": 34.05133056640625, + "grad_norm": 0.05384668707847595, "learning_rate": 1.2592592592592593e-05, - "loss": 0.0105, + "loss": 0.0098, "step": 1500 }, { "epoch": 19.0, "eval_B-Claim": { - "f1-score": 0.6967340590979781, - "precision": 0.7368421052631579, - "recall": 0.6607669616519174, - "support": 339.0 + "f1-score": 0.7074074074074074, + "precision": 0.7100371747211895, + "recall": 0.7047970479704797, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9244712990936558, - "precision": 0.8947368421052632, - "recall": 0.95625, - "support": 160.0 + "f1-score": 0.9310344827586207, + "precision": 0.8940397350993378, + "recall": 0.9712230215827338, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9071840587309911, - "precision": 0.8954451345755694, - "recall": 0.9192348565356004, - "support": 941.0 + "f1-score": 0.8844621513944222, + "precision": 0.8922829581993569, + "recall": 0.8767772511848341, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6769585253456222, - "precision": 0.7378201908588649, - "recall": 0.6253724989357173, - "support": 4698.0 + "f1-score": 0.6726757661601854, + "precision": 0.6937583001328022, + "recall": 0.6528367908022994, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9125809435707678, - "precision": 0.8593205574912892, - "recall": 0.972879684418146, - "support": 2028.0 + "f1-score": 0.9193273214720935, + "precision": 0.9023923444976076, + "recall": 0.936910084451068, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.916641828117238, - "precision": 0.9025042389461327, - "recall": 0.9312293923692887, - "support": 14861.0 + "f1-score": 0.9006964825441325, + "precision": 0.8945445053510833, + "recall": 0.9069336626676077, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9990254872563717, - "precision": 1.0, - "recall": 0.9980528720137797, - "support": 13353.0 + "f1-score": 0.9998674150351351, + "precision": 0.9997348652231551, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9158603628367235, - "eval_loss": 0.69818514585495, + "eval_accuracy": 0.9079952869887224, + "eval_loss": 0.6490303874015808, "eval_macro avg": { - "f1-score": 0.8619423144589463, - "precision": 0.8609527241771824, - "recall": 0.8662551808463499, - "support": 36380.0 - }, - "eval_runtime": 5.0766, - "eval_samples_per_second": 15.956, - "eval_steps_per_second": 2.167, + "f1-score": 0.859353003824571, + "precision": 0.8552556976035047, + "recall": 0.8642111226655746, + "support": 29705.0 + }, + "eval_runtime": 4.8703, + "eval_samples_per_second": 16.426, + "eval_steps_per_second": 2.053, "eval_weighted avg": { - "f1-score": 0.9134424215990907, - "precision": 0.912854812597098, - "recall": 0.9158603628367235, - "support": 36380.0 + "f1-score": 0.9070447031096275, + "precision": 0.9063560601392924, + "recall": 0.9079952869887224, + "support": 29705.0 }, "step": 1539 }, { "epoch": 20.0, "eval_B-Claim": { - "f1-score": 0.7078313253012049, - "precision": 0.7230769230769231, - "recall": 0.6932153392330384, - "support": 339.0 + "f1-score": 0.7035714285714285, + "precision": 0.6816608996539792, + "recall": 0.7269372693726938, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9386503067484663, - "precision": 0.9216867469879518, - "recall": 0.95625, - "support": 160.0 + "f1-score": 0.9059233449477351, + "precision": 0.8783783783783784, + "recall": 0.935251798561151, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9058201058201059, - "precision": 0.9020021074815595, - "recall": 0.9096705632306057, - "support": 941.0 + "f1-score": 0.8838709677419355, + "precision": 0.9028006589785832, + "recall": 0.8657187993680885, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6867509620670699, - "precision": 0.710256993404594, - "recall": 0.6647509578544061, - "support": 4698.0 + "f1-score": 0.6907709278097806, + "precision": 0.6777296777296777, + "recall": 0.7043239190202449, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9304635761589404, - "precision": 0.894090909090909, - "recall": 0.9699211045364892, - "support": 2028.0 + "f1-score": 0.9150943396226415, + "precision": 0.9146401985111663, + "recall": 0.9155489319423745, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9132413977774803, - "precision": 0.9085581085581086, - "recall": 0.9179732184913532, - "support": 14861.0 + "f1-score": 0.9017412935323382, + "precision": 0.9081961345740873, + "recall": 0.8953775582215949, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9989880439263896, - "precision": 0.9999249699879952, - "recall": 0.9980528720137797, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9154205607476635, - "eval_loss": 0.6933491826057434, + "eval_accuracy": 0.9088705605116983, + "eval_loss": 0.6806873083114624, "eval_macro avg": { - "f1-score": 0.8688208168285224, - "precision": 0.8656566797982916, - "recall": 0.8728334364799532, - "support": 36380.0 - }, - "eval_runtime": 5.07, - "eval_samples_per_second": 15.976, - "eval_steps_per_second": 2.17, + "f1-score": 0.857281757460837, + "precision": 0.851915135403696, + "recall": 0.8633083252123068, + "support": 29705.0 + }, + "eval_runtime": 4.8892, + "eval_samples_per_second": 16.362, + "eval_steps_per_second": 2.045, "eval_weighted avg": { - "f1-score": 0.9144315421411552, - "precision": 0.9138389454030841, - "recall": 0.9154205607476635, - "support": 36380.0 + "f1-score": 0.9094792195246973, + "precision": 0.9102297948919554, + "recall": 0.9088705605116983, + "support": 29705.0 }, "step": 1620 }, { "epoch": 21.0, "eval_B-Claim": { - "f1-score": 0.7069219440353463, - "precision": 0.7058823529411765, - "recall": 0.7079646017699115, - "support": 339.0 + "f1-score": 0.7094594594594594, + "precision": 0.6542056074766355, + "recall": 0.7749077490774908, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9153605015673982, - "precision": 0.9182389937106918, - "recall": 0.9125, - "support": 160.0 + "f1-score": 0.9070631970260222, + "precision": 0.9384615384615385, + "recall": 0.8776978417266187, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9070631970260223, - "precision": 0.9065817409766455, - "recall": 0.9075451647183846, - "support": 941.0 + "f1-score": 0.8767346938775511, + "precision": 0.9070945945945946, + "recall": 0.8483412322274881, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6948799473741915, - "precision": 0.7164820257743613, - "recall": 0.6745423584504044, - "support": 4698.0 + "f1-score": 0.6910236955760477, + "precision": 0.648269820411739, + "recall": 0.7398150462384404, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9177033492822966, - "precision": 0.8912639405204461, - "recall": 0.9457593688362919, - "support": 2028.0 + "f1-score": 0.8995290423861854, + "precision": 0.9502487562189055, + "recall": 0.8539493293591655, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9171740148524787, - "precision": 0.9119270937271337, - "recall": 0.9224816634143059, - "support": 14861.0 + "f1-score": 0.8971761541909459, + "precision": 0.9119737561509021, + "recall": 0.8828510938602682, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9989506820566633, - "precision": 0.9997749606181082, - "recall": 0.9981277615517112, - "support": 13353.0 + "f1-score": 0.9999558011049724, + "precision": 0.9999116061168567, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9170973062122045, - "eval_loss": 0.6906760334968567, + "eval_accuracy": 0.9044941928968188, + "eval_loss": 0.7068688869476318, "eval_macro avg": { - "f1-score": 0.8654362337420567, - "precision": 0.8643073011812233, - "recall": 0.866988702677287, - "support": 36380.0 - }, - "eval_runtime": 5.0608, - "eval_samples_per_second": 16.005, - "eval_steps_per_second": 2.174, + "f1-score": 0.8544202919458835, + "precision": 0.858595097061596, + "recall": 0.8539374703556388, + "support": 29705.0 + }, + "eval_runtime": 4.841, + "eval_samples_per_second": 16.525, + "eval_steps_per_second": 2.066, "eval_weighted avg": { - "f1-score": 0.9162839627546457, - "precision": 0.9157494555353419, - "recall": 0.9170973062122045, - "support": 36380.0 + "f1-score": 0.9066064646845915, + "precision": 0.9102050037942175, + "recall": 0.9044941928968188, + "support": 29705.0 }, "step": 1701 }, { "epoch": 22.0, "eval_B-Claim": { - "f1-score": 0.7186147186147187, - "precision": 0.7033898305084746, - "recall": 0.7345132743362832, - "support": 339.0 + "f1-score": 0.7012987012987013, + "precision": 0.7052238805970149, + "recall": 0.6974169741697417, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9325153374233127, - "precision": 0.9156626506024096, - "recall": 0.95, - "support": 160.0 + "f1-score": 0.9310344827586207, + "precision": 0.8940397350993378, + "recall": 0.9712230215827338, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9044038668098818, - "precision": 0.9142236699239956, - "recall": 0.8947927736450585, - "support": 941.0 + "f1-score": 0.8862370723945903, + "precision": 0.8926282051282052, + "recall": 0.8799368088467614, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.697543125980136, - "precision": 0.685432504622971, - "recall": 0.7100893997445722, - "support": 4698.0 + "f1-score": 0.676991150442478, + "precision": 0.7062177572631008, + "recall": 0.6500874781304674, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9283000949667617, - "precision": 0.8951465201465202, - "recall": 0.9640039447731755, - "support": 2028.0 + "f1-score": 0.9239260859131269, + "precision": 0.893686165273909, + "recall": 0.9562841530054644, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9106359835480471, - "precision": 0.9201126528369281, - "recall": 0.9013525334768858, - "support": 14861.0 + "f1-score": 0.904022057858112, + "precision": 0.897072874142274, + "recall": 0.9110797459421313, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9989131656860174, - "precision": 0.999774943735934, - "recall": 0.9980528720137797, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.91412864211105, - "eval_loss": 0.7237672209739685, + "eval_accuracy": 0.9105201144588453, + "eval_loss": 0.6818939447402954, "eval_macro avg": { - "f1-score": 0.8701323275755536, - "precision": 0.8619632531967476, - "recall": 0.8789721139985364, - "support": 36380.0 - }, - "eval_runtime": 5.0999, - "eval_samples_per_second": 15.883, - "eval_steps_per_second": 2.157, + "f1-score": 0.8605013643808042, + "precision": 0.8555526596434059, + "recall": 0.8665754545253285, + "support": 29705.0 + }, + "eval_runtime": 4.8684, + "eval_samples_per_second": 16.432, + "eval_steps_per_second": 2.054, "eval_weighted avg": { - "f1-score": 0.9146496958116109, - "precision": 0.9154631021750581, - "recall": 0.91412864211105, - "support": 36380.0 + "f1-score": 0.9092392748602662, + "precision": 0.9084735380390377, + "recall": 0.9105201144588453, + "support": 29705.0 }, "step": 1782 }, { "epoch": 23.0, "eval_B-Claim": { - "f1-score": 0.6973293768545994, - "precision": 0.7014925373134329, - "recall": 0.6932153392330384, - "support": 339.0 + "f1-score": 0.7009174311926606, + "precision": 0.6970802919708029, + "recall": 0.7047970479704797, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9090909090909091, - "precision": 0.8823529411764706, - "recall": 0.9375, - "support": 160.0 + "f1-score": 0.9181494661921709, + "precision": 0.9084507042253521, + "recall": 0.9280575539568345, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.903672166045769, - "precision": 0.9051172707889126, - "recall": 0.9022316684378321, - "support": 941.0 + "f1-score": 0.885193982581156, + "precision": 0.8873015873015873, + "recall": 0.8830963665086888, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6736772921459087, - "precision": 0.6939742721733243, - "recall": 0.6545338441890166, - "support": 4698.0 + "f1-score": 0.6707700037683707, + "precision": 0.6742424242424242, + "recall": 0.6673331667083229, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9026837806301049, - "precision": 0.8568896765618077, - "recall": 0.9536489151873767, - "support": 2028.0 + "f1-score": 0.9126604580921219, + "precision": 0.925, + "recall": 0.9006458022851466, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9126989456718825, - "precision": 0.9108638831177535, - "recall": 0.9145414171320907, - "support": 14861.0 + "f1-score": 0.8992663532926242, + "precision": 0.895685656777807, + "recall": 0.9028757939308398, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9989130842172332, - "precision": 0.9998499399759904, - "recall": 0.9979779824758481, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9114898295766904, - "eval_loss": 0.7884248495101929, + "eval_accuracy": 0.9058744319138192, + "eval_loss": 0.7413836717605591, "eval_macro avg": { - "f1-score": 0.8568665078080582, - "precision": 0.8500772173010988, - "recall": 0.8648070238078861, - "support": 36380.0 - }, - "eval_runtime": 5.1019, - "eval_samples_per_second": 15.876, - "eval_steps_per_second": 2.156, + "f1-score": 0.8552796707313007, + "precision": 0.8553943806454248, + "recall": 0.8552579616229018, + "support": 29705.0 + }, + "eval_runtime": 4.8821, + "eval_samples_per_second": 16.386, + "eval_steps_per_second": 2.048, "eval_weighted avg": { - "f1-score": 0.9106621435750196, - "precision": 0.9102832766025237, - "recall": 0.9114898295766904, - "support": 36380.0 + "f1-score": 0.9057370428806488, + "precision": 0.905639012166318, + "recall": 0.9058744319138192, + "support": 29705.0 }, "step": 1863 }, { "epoch": 24.0, "eval_B-Claim": { - "f1-score": 0.7227866473149492, - "precision": 0.7114285714285714, - "recall": 0.7345132743362832, - "support": 339.0 + "f1-score": 0.7047970479704797, + "precision": 0.7047970479704797, + "recall": 0.7047970479704797, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9300911854103343, - "precision": 0.9053254437869822, - "recall": 0.95625, - "support": 160.0 + "f1-score": 0.9333333333333332, + "precision": 0.910958904109589, + "recall": 0.9568345323741008, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9065520945220195, - "precision": 0.9163952225841476, - "recall": 0.8969181721572795, - "support": 941.0 + "f1-score": 0.8864177918983319, + "precision": 0.8913738019169329, + "recall": 0.8815165876777251, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.700562572975268, - "precision": 0.6987084480203261, - "recall": 0.70242656449553, - "support": 4698.0 + "f1-score": 0.6843109810408449, + "precision": 0.6969932607568688, + "recall": 0.6720819795051237, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9266207551650438, - "precision": 0.8937242327072835, - "recall": 0.9620315581854043, - "support": 2028.0 + "f1-score": 0.9312638580931263, + "precision": 0.9237536656891495, + "recall": 0.9388971684053651, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9129007142615349, - "precision": 0.9185286103542234, - "recall": 0.9073413632999126, - "support": 14861.0 + "f1-score": 0.9026424370116758, + "precision": 0.8983050847457628, + "recall": 0.9070218772053634, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9998876698992775, - "precision": 0.9998502321401828, - "recall": 0.9999251104620684, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.916245189664651, - "eval_loss": 0.7238907814025879, + "eval_accuracy": 0.9107894293889918, + "eval_loss": 0.7259247303009033, "eval_macro avg": { - "f1-score": 0.8713430913640611, - "precision": 0.8634229658602453, - "recall": 0.8799151489909255, - "support": 36380.0 - }, - "eval_runtime": 5.0787, - "eval_samples_per_second": 15.949, - "eval_steps_per_second": 2.166, + "f1-score": 0.8632522070496845, + "precision": 0.8608831093126833, + "recall": 0.8658784561625941, + "support": 29705.0 + }, + "eval_runtime": 4.8619, + "eval_samples_per_second": 16.454, + "eval_steps_per_second": 2.057, "eval_weighted avg": { - "f1-score": 0.9163125952217944, - "precision": 0.9165644068758583, - "recall": 0.916245189664651, - "support": 36380.0 + "f1-score": 0.910242479146784, + "precision": 0.909787426705373, + "recall": 0.9107894293889918, + "support": 29705.0 }, "step": 1944 }, { "epoch": 24.69, - "grad_norm": 25.948219299316406, + "grad_norm": 0.0542856864631176, "learning_rate": 1.0123456790123458e-05, - "loss": 0.0057, + "loss": 0.004, "step": 2000 }, { "epoch": 25.0, "eval_B-Claim": { - "f1-score": 0.7191679049034176, - "precision": 0.7245508982035929, - "recall": 0.7138643067846607, - "support": 339.0 + "f1-score": 0.7201426024955437, + "precision": 0.696551724137931, + "recall": 0.7453874538745388, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9386503067484663, - "precision": 0.9216867469879518, - "recall": 0.95625, - "support": 160.0 + "f1-score": 0.9285714285714285, + "precision": 0.9219858156028369, + "recall": 0.935251798561151, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9080276448697502, - "precision": 0.9085106382978724, - "recall": 0.9075451647183846, - "support": 941.0 + "f1-score": 0.8883534136546184, + "precision": 0.9035947712418301, + "recall": 0.8736176935229067, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.7102681491170699, - "precision": 0.7278820375335121, - "recall": 0.6934865900383141, - "support": 4698.0 + "f1-score": 0.6924489046628319, + "precision": 0.6784172661870503, + "recall": 0.707073231692077, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9318670810423142, - "precision": 0.9044083526682135, - "recall": 0.9610453648915187, - "support": 2028.0 + "f1-score": 0.9278195488721803, + "precision": 0.936267071320182, + "recall": 0.9195230998509687, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9183714534844725, - "precision": 0.9154242160861136, - "recall": 0.9213377296278851, - "support": 14861.0 + "f1-score": 0.9005723412751232, + "precision": 0.9059180576631259, + "recall": 0.8952893436838391, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9999625538288709, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9999251104620684, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9208356239692138, - "eval_loss": 0.6898869872093201, + "eval_accuracy": 0.9098131627672109, + "eval_loss": 0.7058032751083374, "eval_macro avg": { - "f1-score": 0.8751878705706231, - "precision": 0.8717804128253224, - "recall": 0.8790648952175474, - "support": 36380.0 - }, - "eval_runtime": 5.0592, - "eval_samples_per_second": 16.011, - "eval_steps_per_second": 2.174, + "f1-score": 0.8654154627902466, + "precision": 0.8632478151647079, + "recall": 0.8680203744550686, + "support": 29705.0 + }, + "eval_runtime": 4.8778, + "eval_samples_per_second": 16.401, + "eval_steps_per_second": 2.05, "eval_weighted avg": { - "f1-score": 0.9201627360934556, - "precision": 0.9197045181816026, - "recall": 0.9208356239692138, - "support": 36380.0 + "f1-score": 0.91047415131532, + "precision": 0.9112754497242, + "recall": 0.9098131627672109, + "support": 29705.0 }, "step": 2025 }, { "epoch": 26.0, "eval_B-Claim": { - "f1-score": 0.7195301027900147, - "precision": 0.716374269005848, - "recall": 0.7227138643067846, - "support": 339.0 + "f1-score": 0.7129798903107861, + "precision": 0.7065217391304348, + "recall": 0.7195571955719557, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9226006191950465, - "precision": 0.9141104294478528, - "recall": 0.93125, - "support": 160.0 + "f1-score": 0.9214285714285715, + "precision": 0.9148936170212766, + "recall": 0.9280575539568345, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9099627064464572, - "precision": 0.9123931623931624, - "recall": 0.9075451647183846, - "support": 941.0 + "f1-score": 0.8883610451306412, + "precision": 0.8904761904761904, + "recall": 0.8862559241706162, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.7121407121407122, - "precision": 0.717682663207955, - "recall": 0.7066836951894423, - "support": 4698.0 + "f1-score": 0.6902947154471545, + "precision": 0.7018858176181865, + "recall": 0.6790802299425144, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9204819277108435, - "precision": 0.9000942507068803, - "recall": 0.9418145956607495, - "support": 2028.0 + "f1-score": 0.9235352532274081, + "precision": 0.9230769230769231, + "recall": 0.9239940387481371, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9201467173671636, - "precision": 0.9203015616585891, - "recall": 0.9199919251732723, - "support": 14861.0 + "f1-score": 0.9056852079312161, + "precision": 0.9007853403141362, + "recall": 0.9106386732533521, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9993255395683452, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9986519883172321, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9204233095107202, - "eval_loss": 0.6385065317153931, + "eval_accuracy": 0.9122033327722605, + "eval_loss": 0.6844227313995361, "eval_macro avg": { - "f1-score": 0.8720269036026548, - "precision": 0.8687080480600411, - "recall": 0.8755216047665523, - "support": 36380.0 - }, - "eval_runtime": 5.1139, - "eval_samples_per_second": 15.839, - "eval_steps_per_second": 2.151, + "f1-score": 0.8631835262108254, + "precision": 0.8625199468053067, + "recall": 0.8639405165204871, + "support": 29705.0 + }, + "eval_runtime": 4.8434, + "eval_samples_per_second": 16.517, + "eval_steps_per_second": 2.065, "eval_weighted avg": { - "f1-score": 0.9202441017693053, - "precision": 0.9201301899865153, - "recall": 0.9204233095107202, - "support": 36380.0 + "f1-score": 0.9117462313290554, + "precision": 0.9113620822382424, + "recall": 0.9122033327722605, + "support": 29705.0 }, "step": 2106 }, { "epoch": 27.0, "eval_B-Claim": { - "f1-score": 0.7157894736842104, - "precision": 0.7300613496932515, - "recall": 0.7020648967551623, - "support": 339.0 + "f1-score": 0.6977611940298508, + "precision": 0.7056603773584905, + "recall": 0.6900369003690037, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.914826498422713, - "precision": 0.9235668789808917, - "recall": 0.90625, - "support": 160.0 + "f1-score": 0.9022556390977442, + "precision": 0.9448818897637795, + "recall": 0.8633093525179856, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9131121642969985, - "precision": 0.9050104384133612, - "recall": 0.9213602550478215, - "support": 941.0 + "f1-score": 0.8909657320872274, + "precision": 0.8786482334869432, + "recall": 0.9036334913112164, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.7079230080572964, - "precision": 0.746342614440774, - "recall": 0.6732652192422307, - "support": 4698.0 + "f1-score": 0.6695436118739375, + "precision": 0.7021393307734504, + "recall": 0.6398400399900025, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.922322960058809, - "precision": 0.9167072576716999, - "recall": 0.928007889546351, - "support": 2028.0 + "f1-score": 0.9034411915767849, + "precision": 0.9351408825093036, + "recall": 0.8738201689021361, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9238369972479192, - "precision": 0.9106419139756831, - "recall": 0.9374200928605073, - "support": 14861.0 + "f1-score": 0.9045152378485712, + "precision": 0.8856297548605241, + "recall": 0.9242237120677488, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9998876530726886, + "f1-score": 0.9996904982977407, "precision": 1.0, - "recall": 0.9997753313862053, - "support": 13353.0 + "recall": 0.9993811881188119, + "support": 11312.0 }, - "eval_accuracy": 0.9229246838922485, - "eval_loss": 0.6971738934516907, + "eval_accuracy": 0.9082646019188689, + "eval_loss": 0.7212072610855103, "eval_macro avg": { - "f1-score": 0.8710998221200908, - "precision": 0.876047207596523, - "recall": 0.8668776692626111, - "support": 36380.0 - }, - "eval_runtime": 5.1047, - "eval_samples_per_second": 15.868, - "eval_steps_per_second": 2.155, + "f1-score": 0.8525961578302652, + "precision": 0.8645857812503559, + "recall": 0.8420349790395578, + "support": 29705.0 + }, + "eval_runtime": 4.85, + "eval_samples_per_second": 16.495, + "eval_steps_per_second": 2.062, "eval_weighted avg": { - "f1-score": 0.9215282605927277, - "precision": 0.920789602130938, - "recall": 0.9229246838922485, - "support": 36380.0 + "f1-score": 0.9068521423522986, + "precision": 0.9063105146188606, + "recall": 0.9082646019188689, + "support": 29705.0 }, "step": 2187 }, { "epoch": 28.0, "eval_B-Claim": { - "f1-score": 0.701095461658842, - "precision": 0.7466666666666667, - "recall": 0.6607669616519174, - "support": 339.0 + "f1-score": 0.7015706806282722, + "precision": 0.6655629139072847, + "recall": 0.7416974169741697, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.929663608562691, - "precision": 0.9101796407185628, - "recall": 0.95, - "support": 160.0 + "f1-score": 0.8803088803088803, + "precision": 0.95, + "recall": 0.8201438848920863, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9096605744125327, - "precision": 0.8942505133470225, - "recall": 0.9256110520722636, - "support": 941.0 + "f1-score": 0.886762360446571, + "precision": 0.895330112721417, + "recall": 0.8783570300157978, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6865845152290164, - "precision": 0.7564036885245902, - "recall": 0.6285653469561515, - "support": 4698.0 + "f1-score": 0.675830190974334, + "precision": 0.6582938388625592, + "recall": 0.6943264183954011, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9204761904761904, - "precision": 0.889963167587477, - "recall": 0.953155818540434, - "support": 2028.0 + "f1-score": 0.8754676643506146, + "precision": 0.9473684210526315, + "recall": 0.8137108792846498, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9194842783844231, - "precision": 0.8993115872096764, - "recall": 0.9405827333288473, - "support": 14861.0 + "f1-score": 0.9025816950345251, + "precision": 0.9000087711604245, + "recall": 0.9051693719124911, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9987627938364638, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9975286452482588, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.918938977460143, - "eval_loss": 0.7604945302009583, + "eval_accuracy": 0.9042248779666723, + "eval_loss": 0.7449725866317749, "eval_macro avg": { - "f1-score": 0.8665324889371655, - "precision": 0.8709678948648565, - "recall": 0.8651729368282675, - "support": 36380.0 - }, - "eval_runtime": 5.0746, - "eval_samples_per_second": 15.962, - "eval_steps_per_second": 2.168, + "f1-score": 0.8460744959633139, + "precision": 0.8595091511006167, + "recall": 0.8362007144963709, + "support": 29705.0 + }, + "eval_runtime": 4.8267, + "eval_samples_per_second": 16.574, + "eval_steps_per_second": 2.072, "eval_weighted avg": { - "f1-score": 0.9163178491862402, - "precision": 0.9157871854220297, - "recall": 0.918938977460143, - "support": 36380.0 + "f1-score": 0.9050256946747467, + "precision": 0.9067344629444252, + "recall": 0.9042248779666723, + "support": 29705.0 }, "step": 2268 }, { "epoch": 29.0, "eval_B-Claim": { - "f1-score": 0.7001569858712716, - "precision": 0.7483221476510067, - "recall": 0.6578171091445427, - "support": 339.0 + "f1-score": 0.6972477064220184, + "precision": 0.6934306569343066, + "recall": 0.7011070110701108, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9129129129129129, - "precision": 0.8786127167630058, - "recall": 0.95, - "support": 160.0 + "f1-score": 0.9163636363636365, + "precision": 0.9264705882352942, + "recall": 0.9064748201438849, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9115646258503401, - "precision": 0.8979381443298969, - "recall": 0.9256110520722636, - "support": 941.0 + "f1-score": 0.8855564325177585, + "precision": 0.8848580441640379, + "recall": 0.8862559241706162, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6709571729464077, - "precision": 0.7450623700623701, - "recall": 0.6102596849723286, - "support": 4698.0 + "f1-score": 0.6666666666666666, + "precision": 0.6897381079636558, + "recall": 0.6450887278180455, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9053708439897699, - "precision": 0.8565772107347118, - "recall": 0.9600591715976331, - "support": 2028.0 + "f1-score": 0.9041095890410958, + "precision": 0.9237947122861586, + "recall": 0.8852459016393442, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9177951303087214, - "precision": 0.8991607488702389, - "recall": 0.9372182221923154, - "support": 14861.0 + "f1-score": 0.9035369774919614, + "precision": 0.8903065593423531, + "recall": 0.917166549047283, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9990630036355459, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9981277615517112, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9157778999450248, - "eval_loss": 0.7660219073295593, + "eval_accuracy": 0.9072210065645514, + "eval_loss": 0.7131910920143127, "eval_macro avg": { - "f1-score": 0.8596886679307099, - "precision": 0.86081047691589, - "recall": 0.8627275716472563, - "support": 36380.0 - }, - "eval_runtime": 5.0668, - "eval_samples_per_second": 15.986, - "eval_steps_per_second": 2.171, + "f1-score": 0.8533544297861624, + "precision": 0.8583712384179724, + "recall": 0.8487627048413263, + "support": 29705.0 + }, + "eval_runtime": 4.8395, + "eval_samples_per_second": 16.531, + "eval_steps_per_second": 2.066, "eval_weighted avg": { - "f1-score": 0.9128447549856632, - "precision": 0.912371867054256, - "recall": 0.9157778999450248, - "support": 36380.0 + "f1-score": 0.9062005389193937, + "precision": 0.9055906528951113, + "recall": 0.9072210065645514, + "support": 29705.0 }, "step": 2349 }, { "epoch": 30.0, "eval_B-Claim": { - "f1-score": 0.7267355982274741, - "precision": 0.727810650887574, - "recall": 0.7256637168141593, - "support": 339.0 + "f1-score": 0.707182320441989, + "precision": 0.7058823529411765, + "recall": 0.7084870848708487, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9268292682926829, - "precision": 0.9047619047619048, - "recall": 0.95, - "support": 160.0 + "f1-score": 0.9090909090909091, + "precision": 0.96, + "recall": 0.8633093525179856, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9119999999999999, - "precision": 0.9154175588865097, - "recall": 0.9086078639744952, - "support": 941.0 + "f1-score": 0.8928850664581705, + "precision": 0.8839009287925697, + "recall": 0.9020537124802528, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.7044324324324324, - "precision": 0.7157293497363796, - "recall": 0.6934865900383141, - "support": 4698.0 + "f1-score": 0.6665808776219276, + "precision": 0.6870026525198939, + "recall": 0.6473381654586353, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9097726740098431, - "precision": 0.8669048682447521, - "recall": 0.9571005917159763, - "support": 2028.0 + "f1-score": 0.8990971853425385, + "precision": 0.9657729606389047, + "recall": 0.8410332836562344, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9179300586213867, - "precision": 0.9191687470481075, - "recall": 0.9166947042594711, - "support": 14861.0 + "f1-score": 0.9038552864482148, + "precision": 0.8850934302866322, + "recall": 0.9234297812279464, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9990630036355459, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9981277615517112, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9181693238042881, - "eval_loss": 0.7438024878501892, + "eval_accuracy": 0.9071200134657466, + "eval_loss": 0.784762442111969, "eval_macro avg": { - "f1-score": 0.8709661478884808, - "precision": 0.8642561542236039, - "recall": 0.8785258897648754, - "support": 36380.0 - }, - "eval_runtime": 5.0797, - "eval_samples_per_second": 15.946, - "eval_steps_per_second": 2.165, + "f1-score": 0.8540988064862499, + "precision": 0.8696646178827396, + "recall": 0.8408073400302719, + "support": 29705.0 + }, + "eval_runtime": 4.8562, + "eval_samples_per_second": 16.474, + "eval_steps_per_second": 2.059, "eval_weighted avg": { - "f1-score": 0.9177882762480303, - "precision": 0.9177087395840788, - "recall": 0.9181693238042881, - "support": 36380.0 + "f1-score": 0.906183556468886, + "precision": 0.9063275917791694, + "recall": 0.9071200134657466, + "support": 29705.0 }, "step": 2430 }, { "epoch": 30.86, - "grad_norm": 0.005249341484159231, + "grad_norm": 0.020902352407574654, "learning_rate": 7.654320987654322e-06, - "loss": 0.0029, + "loss": 0.0031, "step": 2500 }, { "epoch": 31.0, "eval_B-Claim": { - "f1-score": 0.7037037037037037, - "precision": 0.7378640776699029, - "recall": 0.672566371681416, - "support": 339.0 + "f1-score": 0.696461824953445, + "precision": 0.7030075187969925, + "recall": 0.6900369003690037, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9102564102564101, - "precision": 0.9342105263157895, - "recall": 0.8875, - "support": 160.0 + "f1-score": 0.8897058823529412, + "precision": 0.9097744360902256, + "recall": 0.8705035971223022, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9135416666666667, - "precision": 0.8958120531154239, - "recall": 0.9319872476089267, - "support": 941.0 + "f1-score": 0.894283476898982, + "precision": 0.8866459627329193, + "recall": 0.9020537124802528, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6881058273463337, - "precision": 0.7410955539179562, - "recall": 0.6421881651766709, - "support": 4698.0 + "f1-score": 0.6603250099088387, + "precision": 0.7003923766816144, + "recall": 0.6245938515371158, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9200885173346448, - "precision": 0.9176066699362433, - "recall": 0.9225838264299803, - "support": 2028.0 + "f1-score": 0.898928024502297, + "precision": 0.9244094488188976, + "recall": 0.8748137108792846, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9197734158872348, - "precision": 0.9007288911823518, - "recall": 0.9396406702106184, - "support": 14861.0 + "f1-score": 0.9050101236376169, + "precision": 0.8843984171086975, + "recall": 0.926605504587156, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9990254872563717, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9980528720137797, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9188015393073117, - "eval_loss": 0.7646034359931946, + "eval_accuracy": 0.9074229927621612, + "eval_loss": 0.7240239381790161, "eval_macro avg": { - "f1-score": 0.864927861207338, - "precision": 0.8753311103053811, - "recall": 0.8563598790173417, - "support": 36380.0 - }, - "eval_runtime": 5.0935, - "eval_samples_per_second": 15.903, - "eval_steps_per_second": 2.16, + "f1-score": 0.8492449060363031, + "precision": 0.8583754514613353, + "recall": 0.841229610996445, + "support": 29705.0 + }, + "eval_runtime": 4.8477, + "eval_samples_per_second": 16.503, + "eval_steps_per_second": 2.063, "eval_weighted avg": { - "f1-score": 0.9167465652664151, - "precision": 0.9159943386289288, - "recall": 0.9188015393073117, - "support": 36380.0 + "f1-score": 0.9056114779535739, + "precision": 0.9048600287675556, + "recall": 0.9074229927621612, + "support": 29705.0 }, "step": 2511 }, { "epoch": 32.0, "eval_B-Claim": { - "f1-score": 0.7127819548872181, - "precision": 0.7269938650306749, - "recall": 0.6991150442477876, - "support": 339.0 + "f1-score": 0.7124773960216998, + "precision": 0.6985815602836879, + "recall": 0.7269372693726938, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.916923076923077, - "precision": 0.9030303030303031, - "recall": 0.93125, - "support": 160.0 + "f1-score": 0.9, + "precision": 0.8936170212765957, + "recall": 0.9064748201438849, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.9111111111111111, - "precision": 0.9072708113804004, - "recall": 0.9149840595111584, - "support": 941.0 + "f1-score": 0.8906624102154829, + "precision": 0.9, + "recall": 0.8815165876777251, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.6915867240048518, - "precision": 0.7174559597346145, - "recall": 0.6675180928054492, - "support": 4698.0 + "f1-score": 0.6955112219451373, + "precision": 0.6939537198308037, + "recall": 0.6970757310672332, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.9048414023372287, - "precision": 0.8762124711316397, - "recall": 0.935404339250493, - "support": 2028.0 + "f1-score": 0.8990129081245255, + "precision": 0.9164086687306502, + "recall": 0.8822652757078987, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.9185957176737817, - "precision": 0.9120456354470682, - "recall": 0.9252405625462621, - "support": 14861.0 + "f1-score": 0.9081789783976418, + "precision": 0.9059071359606776, + "recall": 0.9104622441778405, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9989880439263896, - "precision": 0.9999249699879952, - "recall": 0.9980528720137797, - "support": 13353.0 + "f1-score": 1.0, + "precision": 1.0, + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9169048927982408, - "eval_loss": 0.7867851853370667, + "eval_accuracy": 0.9115973741794311, + "eval_loss": 0.7000880241394043, "eval_macro avg": { - "f1-score": 0.8649754329805226, - "precision": 0.8632762879632423, - "recall": 0.8673664243392756, - "support": 36380.0 - }, - "eval_runtime": 5.0786, - "eval_samples_per_second": 15.949, - "eval_steps_per_second": 2.166, + "f1-score": 0.8579775592434981, + "precision": 0.8583525865832021, + "recall": 0.8578188468781823, + "support": 29705.0 + }, + "eval_runtime": 4.8892, + "eval_samples_per_second": 16.363, + "eval_steps_per_second": 2.045, "eval_weighted avg": { - "f1-score": 0.9159023483997761, - "precision": 0.9152873255952713, - "recall": 0.9169048927982408, - "support": 36380.0 + "f1-score": 0.9116829140602968, + "precision": 0.9118273403549036, + "recall": 0.9115973741794311, + "support": 29705.0 }, "step": 2592 }, { "epoch": 33.0, "eval_B-Claim": { - "f1-score": 0.7119565217391305, - "precision": 0.6599496221662469, - "recall": 0.7728613569321534, - "support": 339.0 + "f1-score": 0.7093235831809871, + "precision": 0.7028985507246377, + "recall": 0.7158671586715867, + "support": 271.0 }, "eval_B-MajorClaim": { - "f1-score": 0.9153605015673982, - "precision": 0.9182389937106918, - "recall": 0.9125, - "support": 160.0 + "f1-score": 0.9154929577464789, + "precision": 0.896551724137931, + "recall": 0.935251798561151, + "support": 139.0 }, "eval_B-Premise": { - "f1-score": 0.8976464148877943, - "precision": 0.9255079006772009, - "recall": 0.871413390010627, - "support": 941.0 + "f1-score": 0.8892430278884461, + "precision": 0.8971061093247589, + "recall": 0.8815165876777251, + "support": 633.0 }, "eval_I-Claim": { - "f1-score": 0.694591622513246, - "precision": 0.6548539114043356, - "recall": 0.7394636015325671, - "support": 4698.0 + "f1-score": 0.6861111111111111, + "precision": 0.693289104363358, + "recall": 0.6790802299425144, + "support": 4001.0 }, "eval_I-MajorClaim": { - "f1-score": 0.913297613248904, - "precision": 0.9023099133782483, - "recall": 0.9245562130177515, - "support": 2028.0 + "f1-score": 0.9068664169787766, + "precision": 0.9116465863453815, + "recall": 0.9021361152508693, + "support": 2013.0 }, "eval_I-Premise": { - "f1-score": 0.906384295389693, - "precision": 0.9266132433572333, - "recall": 0.8870197160352601, - "support": 14861.0 + "f1-score": 0.9057299670691549, + "precision": 0.9016522423288749, + "recall": 0.9098447424135497, + "support": 11336.0 }, "eval_O": { - "f1-score": 0.9991005172026085, + "f1-score": 1.0, "precision": 1.0, - "recall": 0.9982026510896428, - "support": 13353.0 + "recall": 1.0, + "support": 11312.0 }, - "eval_accuracy": 0.9095107201759208, - "eval_loss": 0.8171125650405884, + "eval_accuracy": 0.9103181282612355, + "eval_loss": 0.7453714609146118, "eval_macro avg": { - "f1-score": 0.8626196409355392, - "precision": 0.8553533692419938, - "recall": 0.8722881326597145, - "support": 36380.0 - }, - "eval_runtime": 5.0657, - "eval_samples_per_second": 15.99, - "eval_steps_per_second": 2.171, + "f1-score": 0.8589667234249935, + "precision": 0.8575920453178488, + "recall": 0.8605280903596281, + "support": 29705.0 + }, + "eval_runtime": 4.8419, + "eval_samples_per_second": 16.523, + "eval_steps_per_second": 2.065, "eval_weighted avg": { - "f1-score": 0.9114519362220568, - "precision": 0.9145500738066468, - "recall": 0.9095107201759208, - "support": 36380.0 + "f1-score": 0.91002789457842, + "precision": 0.9097828132642751, + "recall": 0.9103181282612355, + "support": 29705.0 }, "step": 2673 } @@ -2128,7 +2128,7 @@ "num_input_tokens_seen": 0, "num_train_epochs": 50, "save_steps": 500, - "total_flos": 4730360494167000.0, + "total_flos": 4745096819694000.0, "train_batch_size": 4, "trial_name": null, "trial_params": null