diff --git "a/checkpoint-3888/trainer_state.json" "b/checkpoint-3888/trainer_state.json" --- "a/checkpoint-3888/trainer_state.json" +++ "b/checkpoint-3888/trainer_state.json" @@ -11,2209 +11,2209 @@ { "epoch": 1.0, "eval_Claim": { - "f1-score": 0.5724454649827784, - "precision": 0.6002889477486154, - "recall": 0.5470704410796576, - "support": 4557.0 + "f1-score": 0.5385059461955894, + "precision": 0.6267132491265789, + "recall": 0.4720647773279352, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.7860447185813415, - "precision": 0.698526892771497, - "recall": 0.8986337593653592, - "support": 2269.0 + "f1-score": 0.7814784727863526, + "precision": 0.9588039867109635, + "recall": 0.6595063985374772, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9987820830889982, - "precision": 0.9975671292124707, - "recall": 1.0, - "support": 11071.0 + "f1-score": 0.9938608715302625, + "precision": 0.9996969237763298, + "recall": 0.9880925634688834, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.8965445389824622, - "precision": 0.9051258677512096, - "recall": 0.8881243979633962, - "support": 14534.0 + "f1-score": 0.8968837690149166, + "precision": 0.8455112497215416, + "recall": 0.954902824076986, + "support": 15899.0 }, - "eval_accuracy": 0.8791279948197712, - "eval_loss": 0.3157936632633209, + "eval_accuracy": 0.8837548103353491, + "eval_loss": 0.2860471308231354, "eval_macro avg": { - "f1-score": 0.8134542014088951, - "precision": 0.8003772093709481, - "recall": 0.8334571496021033, - "support": 32431.0 - }, - "eval_runtime": 4.8184, - "eval_samples_per_second": 16.603, - "eval_steps_per_second": 2.075, + "f1-score": 0.8026822648817803, + "precision": 0.8576813523338533, + "recall": 0.7686416408528204, + "support": 36380.0 + }, + "eval_runtime": 4.9572, + "eval_samples_per_second": 16.34, + "eval_steps_per_second": 2.219, "eval_weighted avg": { - "f1-score": 0.8781740995293698, - "precision": 0.8793943851252293, - "recall": 0.8791279948197712, - "support": 32431.0 + "f1-score": 0.8768739289107891, + "precision": 0.8792073379911217, + "recall": 0.8837548103353491, + "support": 36380.0 }, "step": 81 }, { "epoch": 2.0, "eval_Claim": { - "f1-score": 0.6090644692195831, - "precision": 0.680108254397835, - "recall": 0.5514592933947773, - "support": 4557.0 + "f1-score": 0.6469459868753155, + "precision": 0.6453172205438067, + "recall": 0.648582995951417, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8585858585858586, - "precision": 0.8960229995208433, - "recall": 0.8241516086381666, - "support": 2269.0 + "f1-score": 0.8435970110796187, + "precision": 0.9669226225634967, + "recall": 0.7481718464351006, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.999458141425088, - "precision": 0.9992776523702032, - "recall": 0.9996386956914461, - "support": 11071.0 + "f1-score": 0.998877245508982, + "precision": 0.9983541557567143, + "recall": 0.9994008836965476, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9105885478942475, - "precision": 0.8801849235905997, - "recall": 0.9431677445988716, - "support": 14534.0 + "f1-score": 0.9099026477336144, + "precision": 0.8972179761540813, + "recall": 0.922951129001824, + "support": 15899.0 }, - "eval_accuracy": 0.8990780426135487, - "eval_loss": 0.24451124668121338, + "eval_accuracy": 0.903243540406817, + "eval_loss": 0.24885737895965576, "eval_macro avg": { - "f1-score": 0.8444242542811943, - "precision": 0.8638984574698704, - "recall": 0.8296043355808155, - "support": 32431.0 - }, - "eval_runtime": 4.8529, - "eval_samples_per_second": 16.485, - "eval_steps_per_second": 2.061, + "f1-score": 0.8498307227993828, + "precision": 0.8769529937545246, + "recall": 0.8297767137712222, + "support": 36380.0 + }, + "eval_runtime": 5.0097, + "eval_samples_per_second": 16.169, + "eval_steps_per_second": 2.196, "eval_weighted avg": { - "f1-score": 0.8949194640367888, - "precision": 0.8938342965082798, - "recall": 0.8990780426135487, - "support": 32431.0 + "f1-score": 0.9028656814459712, + "precision": 0.9043261520711511, + "recall": 0.903243540406817, + "support": 36380.0 }, "step": 162 }, { "epoch": 3.0, "eval_Claim": { - "f1-score": 0.6542320859402667, - "precision": 0.7079713847726111, - "recall": 0.60807548825982, - "support": 4557.0 + "f1-score": 0.7040630102775911, + "precision": 0.6698958142935478, + "recall": 0.7419028340080972, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8915187376725839, - "precision": 0.8866608544027899, - "recall": 0.8964301454385192, - "support": 2269.0 + "f1-score": 0.8888411094388303, + "precision": 0.8392204628501827, + "recall": 0.9446983546617916, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9995936611133684, - "precision": 0.9992778479870013, - "recall": 0.9999096739228616, - "support": 11071.0 + "f1-score": 0.998989104796136, + "precision": 0.9988769092542678, + "recall": 0.9991013255448213, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9157990498332153, - "precision": 0.8973258501155497, - "recall": 0.935048850970139, - "support": 14534.0 + "f1-score": 0.914582593823615, + "precision": 0.9390987408880053, + "recall": 0.8913139191144097, + "support": 15899.0 }, - "eval_accuracy": 0.9085442940396534, - "eval_loss": 0.22665414214134216, + "eval_accuracy": 0.9137987905442551, + "eval_loss": 0.24298855662345886, "eval_macro avg": { - "f1-score": 0.8652858836398586, - "precision": 0.872808984319488, - "recall": 0.8598660396478349, - "support": 32431.0 - }, - "eval_runtime": 4.8494, - "eval_samples_per_second": 16.497, - "eval_steps_per_second": 2.062, + "f1-score": 0.8766189545840432, + "precision": 0.8617729818215009, + "recall": 0.8942541083322799, + "support": 36380.0 + }, + "eval_runtime": 4.9876, + "eval_samples_per_second": 16.24, + "eval_steps_per_second": 2.205, "eval_weighted avg": { - "f1-score": 0.9059516032151627, - "precision": 0.9047762338408377, - "recall": 0.9085442940396534, - "support": 32431.0 + "f1-score": 0.9154289662937856, + "precision": 0.9184781183611038, + "recall": 0.9137987905442551, + "support": 36380.0 }, "step": 243 }, { "epoch": 4.0, "eval_Claim": { - "f1-score": 0.6861642294713161, - "precision": 0.6424741478360781, - "recall": 0.7362299758613122, - "support": 4557.0 + "f1-score": 0.6981611893583725, + "precision": 0.6754352763058289, + "recall": 0.7224696356275304, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8995515695067264, - "precision": 0.9155636695572797, - "recall": 0.884089907448215, - "support": 2269.0 + "f1-score": 0.901231802911534, + "precision": 0.8836187966622749, + "recall": 0.9195612431444241, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9976481230212574, - "precision": 0.9990941208442794, - "recall": 0.9962063047601842, - "support": 11071.0 + "f1-score": 0.9989130842172332, + "precision": 0.9998499399759904, + "recall": 0.9979779824758481, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9062532879739067, - "precision": 0.9242435081193219, - "recall": 0.8889500481629283, - "support": 14534.0 + "f1-score": 0.9146224912392481, + "precision": 0.9266670970240785, + "recall": 0.9028869740235235, + "support": 15899.0 }, - "eval_accuracy": 0.9037649162838025, - "eval_loss": 0.2452327460050583, + "eval_accuracy": 0.9142935678944475, + "eval_loss": 0.23980534076690674, "eval_macro avg": { - "f1-score": 0.8724043024933017, - "precision": 0.8703438615892398, - "recall": 0.8763690590581599, - "support": 32431.0 - }, - "eval_runtime": 4.8393, - "eval_samples_per_second": 16.531, - "eval_steps_per_second": 2.066, + "f1-score": 0.8782321419315969, + "precision": 0.8713927774920431, + "recall": 0.8857239588178316, + "support": 36380.0 + }, + "eval_runtime": 4.9726, + "eval_samples_per_second": 16.289, + "eval_steps_per_second": 2.212, "eval_weighted avg": { - "f1-score": 0.9060584182508294, - "precision": 0.9095955973231696, - "recall": 0.9037649162838025, - "support": 32431.0 + "f1-score": 0.9153623106642749, + "precision": 0.9168247558035478, + "recall": 0.9142935678944475, + "support": 36380.0 }, "step": 324 }, { "epoch": 5.0, "eval_Claim": { - "f1-score": 0.6640249759846301, - "precision": 0.7332272606735614, - "recall": 0.6067588325652842, - "support": 4557.0 + "f1-score": 0.6650318783717509, + "precision": 0.6450999048525214, + "recall": 0.6862348178137652, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8848007008322383, - "precision": 0.8794079233783195, - "recall": 0.8902600264433671, - "support": 2269.0 + "f1-score": 0.9172031076581575, + "precision": 0.8916702632714717, + "recall": 0.9442413162705667, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.999367774566474, - "precision": 0.9992775219001174, - "recall": 0.9994580435371692, - "support": 11071.0 + "f1-score": 0.9982383147794145, + "precision": 0.9992495872730002, + "recall": 0.9972290870965326, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9201314377682404, - "precision": 0.8973839110529758, - "recall": 0.9440621989816981, - "support": 14534.0 + "f1-score": 0.8998438545616774, + "precision": 0.9119622787753521, + "recall": 0.8880432731618341, + "support": 15899.0 }, - "eval_accuracy": 0.9118127717307515, - "eval_loss": 0.28411003947257996, + "eval_accuracy": 0.9040956569543706, + "eval_loss": 0.2932502031326294, "eval_macro avg": { - "f1-score": 0.8670812222878956, - "precision": 0.8773241542512435, - "recall": 0.8601347753818797, - "support": 32431.0 - }, - "eval_runtime": 4.8228, - "eval_samples_per_second": 16.588, - "eval_steps_per_second": 2.073, + "f1-score": 0.87007928884275, + "precision": 0.8619955085430864, + "recall": 0.8789371235856747, + "support": 36380.0 + }, + "eval_runtime": 4.9595, + "eval_samples_per_second": 16.332, + "eval_steps_per_second": 2.218, "eval_weighted avg": { - "f1-score": 0.9087220731552942, - "precision": 0.9078434958291443, - "recall": 0.9118127717307515, - "support": 32431.0 + "f1-score": 0.9051179642561333, + "precision": 0.9065430476942037, + "recall": 0.9040956569543706, + "support": 36380.0 }, "step": 405 }, { "epoch": 6.0, "eval_Claim": { - "f1-score": 0.6496087442553721, - "precision": 0.7484258729250143, - "recall": 0.5738424402018872, - "support": 4557.0 + "f1-score": 0.6622698072805139, + "precision": 0.5740163325909429, + "recall": 0.7825910931174089, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8865800865800866, - "precision": 0.8711186729051468, - "recall": 0.9026002644336713, - "support": 2269.0 + "f1-score": 0.9223704351629127, + "precision": 0.8842767295597485, + "recall": 0.9638939670932358, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9986892655367232, - "precision": 0.9994572100597069, - "recall": 0.9979225002258152, - "support": 11071.0 + "f1-score": 0.9982004948639124, + "precision": 0.9993994444861497, + "recall": 0.9970044184827379, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9191778088205947, - "precision": 0.889647180015452, - "recall": 0.9507362047612495, - "support": 14534.0 + "f1-score": 0.8780749379985253, + "precision": 0.9398091685199799, + "recall": 0.8239511918988616, + "support": 15899.0 }, - "eval_accuracy": 0.9105177145323918, - "eval_loss": 0.3082066774368286, + "eval_accuracy": 0.8902693787795491, + "eval_loss": 0.3604726195335388, "eval_macro avg": { - "f1-score": 0.8635139762981943, - "precision": 0.87716223397633, - "recall": 0.8562753524056559, - "support": 32431.0 - }, - "eval_runtime": 4.8389, - "eval_samples_per_second": 16.533, - "eval_steps_per_second": 2.067, + "f1-score": 0.865228918826466, + "precision": 0.8493754187892053, + "recall": 0.8918601676480611, + "support": 36380.0 + }, + "eval_runtime": 4.991, + "eval_samples_per_second": 16.229, + "eval_steps_per_second": 2.204, "eval_weighted avg": { - "f1-score": 0.9061618943658084, - "precision": 0.9059932736781678, - "recall": 0.9105177145323918, - "support": 32431.0 + "f1-score": 0.8955262233990811, + "precision": 0.9086708334194257, + "recall": 0.8902693787795491, + "support": 36380.0 }, "step": 486 }, { "epoch": 6.17, - "grad_norm": 0.8612700700759888, + "grad_norm": 17.04454231262207, "learning_rate": 1.7530864197530865e-05, - "loss": 0.2223, + "loss": 0.2104, "step": 500 }, { "epoch": 7.0, "eval_Claim": { - "f1-score": 0.6977381768334475, - "precision": 0.629950495049505, - "recall": 0.7818740399385561, - "support": 4557.0 + "f1-score": 0.7037529399734124, + "precision": 0.7110973341599504, + "recall": 0.6965587044534413, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8892209178228387, - "precision": 0.8621688741721855, - "recall": 0.9180255619215514, - "support": 2269.0 + "f1-score": 0.9196885428253615, + "precision": 0.8959687906371911, + "recall": 0.9446983546617916, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9988698521766648, - "precision": 0.9998190045248869, - "recall": 0.9979225002258152, - "support": 11071.0 + "f1-score": 0.9987257327036954, + "precision": 0.9996248780853777, + "recall": 0.997828203399985, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9034227633516503, - "precision": 0.9449996243143738, - "recall": 0.8653502132929682, - "support": 14534.0 + "f1-score": 0.9188781285372909, + "precision": 0.9187048098082364, + "recall": 0.9190515126737531, + "support": 15899.0 }, - "eval_accuracy": 0.9025623631710401, - "eval_loss": 0.3512111008167267, + "eval_accuracy": 0.9192963166575041, + "eval_loss": 0.36414769291877747, "eval_macro avg": { - "f1-score": 0.8723129275461503, - "precision": 0.8592344995152377, - "recall": 0.8907930788447227, - "support": 32431.0 - }, - "eval_runtime": 4.8174, - "eval_samples_per_second": 16.607, - "eval_steps_per_second": 2.076, + "f1-score": 0.8852613360099401, + "precision": 0.8813489531726889, + "recall": 0.8895341937972427, + "support": 36380.0 + }, + "eval_runtime": 4.9835, + "eval_samples_per_second": 16.254, + "eval_steps_per_second": 2.207, "eval_weighted avg": { - "f1-score": 0.9061105026169647, - "precision": 0.9136494810618672, - "recall": 0.9025623631710401, - "support": 32431.0 + "f1-score": 0.9190227083446231, + "precision": 0.9188476996393494, + "recall": 0.9192963166575041, + "support": 36380.0 }, "step": 567 }, { "epoch": 8.0, "eval_Claim": { - "f1-score": 0.6987032799389779, - "precision": 0.6177710335525206, - "recall": 0.80403774412991, - "support": 4557.0 + "f1-score": 0.7025399811853246, + "precision": 0.6562390158172232, + "recall": 0.7558704453441295, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9015364639688379, - "precision": 0.8856292517006803, - "recall": 0.9180255619215514, - "support": 2269.0 + "f1-score": 0.9105835367206567, + "precision": 0.8848641655886158, + "recall": 0.9378427787934186, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.998915499322187, - "precision": 0.9994574554661362, - "recall": 0.9983741306115076, - "support": 11071.0 + "f1-score": 0.9986513823331086, + "precision": 0.9991005172026085, + "recall": 0.9982026510896428, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.8976577489773016, - "precision": 0.9472075788830315, - "recall": 0.8530342644832806, - "support": 14534.0 + "f1-score": 0.9083384525849526, + "precision": 0.9345974717232203, + "recall": 0.8835146864582678, + "support": 15899.0 }, - "eval_accuracy": 0.9003114304215103, - "eval_loss": 0.3705728054046631, + "eval_accuracy": 0.911544804837823, + "eval_loss": 0.35222452878952026, "eval_macro avg": { - "f1-score": 0.8742032480518261, - "precision": 0.8625163299005921, - "recall": 0.8933679252865624, - "support": 32431.0 - }, - "eval_runtime": 4.8542, - "eval_samples_per_second": 16.48, - "eval_steps_per_second": 2.06, + "f1-score": 0.8800283382060107, + "precision": 0.8687002925829169, + "recall": 0.8938576404213647, + "support": 36380.0 + }, + "eval_runtime": 4.9741, + "eval_samples_per_second": 16.285, + "eval_steps_per_second": 2.211, "eval_weighted avg": { - "f1-score": 0.9045397397569994, - "precision": 0.9144455555782818, - "recall": 0.9003114304215103, - "support": 32431.0 + "f1-score": 0.9136769997620249, + "precision": 0.9174837257168467, + "recall": 0.911544804837823, + "support": 36380.0 }, "step": 648 }, { "epoch": 9.0, "eval_Claim": { - "f1-score": 0.6657139454567107, - "precision": 0.7278645833333334, - "recall": 0.6133421110379635, - "support": 4557.0 + "f1-score": 0.6566933693658913, + "precision": 0.6275594908688434, + "recall": 0.6886639676113361, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8882521489971346, - "precision": 0.8884479717813051, - "recall": 0.8880564125165271, - "support": 2269.0 + "f1-score": 0.8702745722244328, + "precision": 0.7706131078224101, + "recall": 0.9995429616087751, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9987339482727438, - "precision": 0.9999094612947035, - "recall": 0.9975611959172613, - "support": 11071.0 + "f1-score": 0.9987648313807688, + "precision": 0.9983537862915295, + "recall": 0.9991762150827529, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9178854152690192, - "precision": 0.8955360649299646, - "recall": 0.9413788358332187, - "support": 14534.0 + "f1-score": 0.9001826722338205, + "precision": 0.9350138917124077, + "recall": 0.8678533241084345, + "support": 15899.0 }, - "eval_accuracy": 0.9107335573987851, - "eval_loss": 0.41826051473617554, + "eval_accuracy": 0.8996426608026388, + "eval_loss": 0.4436376690864563, "eval_macro avg": { - "f1-score": 0.8676463644989021, - "precision": 0.8779395203348266, - "recall": 0.8600846388262426, - "support": 32431.0 - }, - "eval_runtime": 4.809, - "eval_samples_per_second": 16.635, - "eval_steps_per_second": 2.079, + "f1-score": 0.8564788613012284, + "precision": 0.8328850691737977, + "recall": 0.8888091171028246, + "support": 36380.0 + }, + "eval_runtime": 5.0402, + "eval_samples_per_second": 16.071, + "eval_steps_per_second": 2.182, "eval_weighted avg": { - "f1-score": 0.9079779452489347, - "precision": 0.9071100542045434, - "recall": 0.9107335573987851, - "support": 32431.0 + "f1-score": 0.9015045934020747, + "precision": 0.90662587513727, + "recall": 0.8996426608026388, + "support": 36380.0 }, "step": 729 }, { "epoch": 10.0, "eval_Claim": { - "f1-score": 0.6738416222952305, - "precision": 0.6964871194379391, - "recall": 0.652622339258284, - "support": 4557.0 + "f1-score": 0.7371103705155853, + "precision": 0.7145572025845686, + "recall": 0.7611336032388664, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8951019550136641, - "precision": 0.8557073954983923, - "recall": 0.9382988100484795, - "support": 2269.0 + "f1-score": 0.9365328549002018, + "precision": 0.9194187582562747, + "recall": 0.9542961608775137, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9989605459393501, - "precision": 0.9996382054992764, - "recall": 0.998283804534369, - "support": 11071.0 + "f1-score": 0.9991386090408599, + "precision": 0.9993257416841475, + "recall": 0.9989515464689583, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9109121470961545, - "precision": 0.9083259218717931, - "recall": 0.9135131415990092, - "support": 14534.0 + "f1-score": 0.9225428371233836, + "precision": 0.9344473837021743, + "recall": 0.9109377948298635, + "support": 15899.0 }, - "eval_accuracy": 0.9075267490980852, - "eval_loss": 0.5459077954292297, + "eval_accuracy": 0.9255085211654756, + "eval_loss": 0.44094258546829224, "eval_macro avg": { - "f1-score": 0.8697040675860997, - "precision": 0.8650396605768502, - "recall": 0.8756795238600354, - "support": 32431.0 - }, - "eval_runtime": 4.8478, - "eval_samples_per_second": 16.502, - "eval_steps_per_second": 2.063, + "f1-score": 0.8988311678950076, + "precision": 0.8919372715567913, + "recall": 0.9063297763538004, + "support": 36380.0 + }, + "eval_runtime": 4.9458, + "eval_samples_per_second": 16.378, + "eval_steps_per_second": 2.224, "eval_weighted avg": { - "f1-score": 0.9065515080853328, - "precision": 0.9060496255197702, - "recall": 0.9075267490980852, - "support": 32431.0 + "f1-score": 0.9263184587909816, + "precision": 0.927497977075915, + "recall": 0.9255085211654756, + "support": 36380.0 }, "step": 810 }, { "epoch": 11.0, "eval_Claim": { - "f1-score": 0.6581417175035196, - "precision": 0.7070834383665239, - "recall": 0.6155365371955234, - "support": 4557.0 + "f1-score": 0.6750768578394996, + "precision": 0.7086579123080348, + "recall": 0.6445344129554655, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.881979417560762, - "precision": 0.876414273281114, - "recall": 0.8876156897311591, - "support": 2269.0 + "f1-score": 0.9008264462809917, + "precision": 0.8220211161387632, + "recall": 0.9963436928702011, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548390010388, - "precision": 0.9999096820809249, - "recall": 1.0, - "support": 11071.0 + "f1-score": 0.9990631440884392, + "precision": 0.9998499849984999, + "recall": 0.9982775406275743, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9173079519373565, - "precision": 0.9002915065588976, - "recall": 0.9349800467868447, - "support": 14534.0 + "f1-score": 0.9168605748066159, + "precision": 0.9167452681883921, + "recall": 0.9169759104346186, + "support": 15899.0 }, - "eval_accuracy": 0.9089759797724399, - "eval_loss": 0.5726844668388367, + "eval_accuracy": 0.9145959318306762, + "eval_loss": 0.4954843521118164, "eval_macro avg": { - "f1-score": 0.8643459815006692, - "precision": 0.8709247250718651, - "recall": 0.8595330684283817, - "support": 32431.0 - }, - "eval_runtime": 4.8447, - "eval_samples_per_second": 16.513, - "eval_steps_per_second": 2.064, + "f1-score": 0.8729567557538865, + "precision": 0.8618185704084225, + "recall": 0.8890328892219649, + "support": 36380.0 + }, + "eval_runtime": 5.0503, + "eval_samples_per_second": 16.039, + "eval_steps_per_second": 2.178, "eval_weighted avg": { - "f1-score": 0.9066330640790278, - "precision": 0.9054793272287636, - "recall": 0.9089759797724399, - "support": 32431.0 + "f1-score": 0.9132365141301059, + "precision": 0.9132952764575458, + "recall": 0.9145959318306762, + "support": 36380.0 }, "step": 891 }, { "epoch": 12.0, "eval_Claim": { - "f1-score": 0.7079179557694512, - "precision": 0.7407673860911271, - "recall": 0.6778582400702217, - "support": 4557.0 + "f1-score": 0.6857200770244248, + "precision": 0.6866247209255125, + "recall": 0.6848178137651821, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8947480331703169, - "precision": 0.8644207066557108, - "recall": 0.9272807404142794, - "support": 2269.0 + "f1-score": 0.9214571488734471, + "precision": 0.8543537680593518, + "recall": 1.0, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.999186477447347, - "precision": 0.9999095431931253, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9987652935234033, + "precision": 0.9979811574697174, + "recall": 0.9995506627724107, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9243840851702722, - "precision": 0.9169374492282697, - "recall": 0.9319526627218935, - "support": 14534.0 + "f1-score": 0.9127542413343095, + "precision": 0.9239592731022039, + "recall": 0.9018177243851815, + "support": 15899.0 }, - "eval_accuracy": 0.9186272393697388, - "eval_loss": 0.5660926699638367, + "eval_accuracy": 0.91412864211105, + "eval_loss": 0.5210011005401611, "eval_macro avg": { - "f1-score": 0.8815591378893468, - "precision": 0.8805087712920582, - "recall": 0.8838890249737601, - "support": 32431.0 - }, - "eval_runtime": 4.8659, - "eval_samples_per_second": 16.441, - "eval_steps_per_second": 2.055, + "f1-score": 0.8796741901888963, + "precision": 0.8657297298891964, + "recall": 0.8965465502306935, + "support": 36380.0 + }, + "eval_runtime": 4.9497, + "eval_samples_per_second": 16.365, + "eval_steps_per_second": 2.222, "eval_weighted avg": { - "f1-score": 0.917429533390557, - "precision": 0.9168331226787283, - "recall": 0.9186272393697388, - "support": 32431.0 + "f1-score": 0.9140186385274314, + "precision": 0.9147147620844138, + "recall": 0.91412864211105, + "support": 36380.0 }, "step": 972 }, { "epoch": 12.35, - "grad_norm": 2.9101791381835938, + "grad_norm": 3.2667109966278076, "learning_rate": 1.506172839506173e-05, - "loss": 0.0294, + "loss": 0.0286, "step": 1000 }, { "epoch": 13.0, "eval_Claim": { - "f1-score": 0.6744648496900223, - "precision": 0.7221943887775552, - "recall": 0.6326530612244898, - "support": 4557.0 + "f1-score": 0.7024380291747424, + "precision": 0.7079991774624718, + "recall": 0.6969635627530364, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9044719714465672, - "precision": 0.863672814755413, - "recall": 0.9493168796826796, - "support": 2269.0 + "f1-score": 0.9075119409465915, + "precision": 0.8643507030603805, + "recall": 0.9552102376599635, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9995934041111363, - "precision": 0.9999096167751266, - "recall": 0.9992773913828923, - "support": 11071.0 + "f1-score": 0.9986895802912875, + "precision": 0.9985774183887391, + "recall": 0.9988017673930952, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9161992180860105, - "precision": 0.9055171023452725, - "recall": 0.9271363698912893, - "support": 14534.0 + "f1-score": 0.9200429808482398, + "precision": 0.9246014101505431, + "recall": 0.9155292785709793, + "support": 15899.0 }, - "eval_accuracy": 0.9119361105115475, - "eval_loss": 0.5764271020889282, + "eval_accuracy": 0.9188015393073117, + "eval_loss": 0.5909315347671509, "eval_macro avg": { - "f1-score": 0.8736823608334341, - "precision": 0.8728234806633418, - "recall": 0.8770959255453377, - "support": 32431.0 - }, - "eval_runtime": 4.8358, - "eval_samples_per_second": 16.543, - "eval_steps_per_second": 2.068, + "f1-score": 0.8821706328152152, + "precision": 0.8738821772655336, + "recall": 0.8916262115942686, + "support": 36380.0 + }, + "eval_runtime": 4.9699, + "eval_samples_per_second": 16.298, + "eval_steps_per_second": 2.213, "eval_weighted avg": { - "f1-score": 0.9098800911419986, - "precision": 0.909053016531805, - "recall": 0.9119361105115475, - "support": 32431.0 + "f1-score": 0.9186076225687217, + "precision": 0.918717904416961, + "recall": 0.9188015393073117, + "support": 36380.0 }, "step": 1053 }, { "epoch": 14.0, "eval_Claim": { - "f1-score": 0.7047413793103449, - "precision": 0.6923565530383231, - "recall": 0.717577353522054, - "support": 4557.0 + "f1-score": 0.7095000509113125, + "precision": 0.7137881581643106, + "recall": 0.7052631578947368, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9112001687407719, - "precision": 0.8737864077669902, - "recall": 0.9519612163948876, - "support": 2269.0 + "f1-score": 0.921028466483012, + "precision": 0.9252767527675276, + "recall": 0.916819012797075, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, + "f1-score": 0.9994754982766372, "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "recall": 0.9989515464689583, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9145424153781864, - "precision": 0.9258919757439007, - "recall": 0.903467730838035, - "support": 14534.0 + "f1-score": 0.9206986297074409, + "precision": 0.9180215107553776, + "recall": 0.9233914082646707, + "support": 15899.0 }, - "eval_accuracy": 0.9131694983195091, - "eval_loss": 0.5468704700469971, + "eval_accuracy": 0.9211105002748763, + "eval_loss": 0.559357225894928, "eval_macro avg": { - "f1-score": 0.8824289004618456, - "precision": 0.8730087341373035, - "recall": 0.8928676893609057, - "support": 32431.0 - }, - "eval_runtime": 4.8411, - "eval_samples_per_second": 16.525, - "eval_steps_per_second": 2.066, + "f1-score": 0.8876756613446006, + "precision": 0.889271605421804, + "recall": 0.8861062813563602, + "support": 36380.0 + }, + "eval_runtime": 4.9811, + "eval_samples_per_second": 16.262, + "eval_steps_per_second": 2.208, "eval_weighted avg": { - "f1-score": 0.9137390947680665, - "precision": 0.9147298617643856, - "recall": 0.9131694983195091, - "support": 32431.0 + "f1-score": 0.9209544908843665, + "precision": 0.9208148168193181, + "recall": 0.9211105002748763, + "support": 36380.0 }, "step": 1134 }, { "epoch": 15.0, "eval_Claim": { - "f1-score": 0.6868638486408813, - "precision": 0.7559304164470216, - "recall": 0.6293614219881501, - "support": 4557.0 + "f1-score": 0.6915445526531356, + "precision": 0.6282454109475774, + "recall": 0.7690283400809717, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9058546000428909, - "precision": 0.8822055137844611, - "recall": 0.9308065226972234, - "support": 2269.0 + "f1-score": 0.9327135900841081, + "precision": 0.9042918454935622, + "recall": 0.9629798903107861, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, - "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "f1-score": 0.9995130539011873, + "precision": 0.9998501199040767, + "recall": 0.9991762150827529, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9218029420675262, - "precision": 0.9023924075660713, - "recall": 0.9420668776661622, - "support": 14534.0 + "f1-score": 0.8984226716408141, + "precision": 0.936421311139914, + "recall": 0.8633876344424177, + "support": 15899.0 }, - "eval_accuracy": 0.9170855046097869, - "eval_loss": 0.6552095413208008, + "eval_accuracy": 0.9064046179219352, + "eval_loss": 0.6213898658752441, "eval_macro avg": { - "f1-score": 0.8786190564182342, - "precision": 0.8851320844493885, - "recall": 0.8755361240685993, - "support": 32431.0 - }, - "eval_runtime": 4.8081, - "eval_samples_per_second": 16.639, - "eval_steps_per_second": 2.08, + "f1-score": 0.8805484670698113, + "precision": 0.8672021718712826, + "recall": 0.898643019979232, + "support": 36380.0 + }, + "eval_runtime": 4.9756, + "eval_samples_per_second": 16.28, + "eval_steps_per_second": 2.211, "eval_weighted avg": { - "f1-score": 0.9143537536055225, - "precision": 0.9137205288178378, - "recall": 0.9170855046097869, - "support": 32431.0 + "f1-score": 0.9094977265082566, + "precision": 0.9159231436204942, + "recall": 0.9064046179219352, + "support": 36380.0 }, "step": 1215 }, { "epoch": 16.0, "eval_Claim": { - "f1-score": 0.7261954261954263, - "precision": 0.6899071696622556, - "recall": 0.7665130568356375, - "support": 4557.0 + "f1-score": 0.6819089456869009, + "precision": 0.6727738376674547, + "recall": 0.6912955465587044, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9104052573932091, - "precision": 0.9050522648083623, - "recall": 0.9158219479947113, - "support": 2269.0 + "f1-score": 0.9290586630286495, + "precision": 0.9244343891402715, + "recall": 0.9337294332723949, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9998127410958391, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.9996255523103422, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9186291000841041, - "precision": 0.9360805599200114, - "recall": 0.9018164304389706, - "support": 14534.0 + "f1-score": 0.9087691578448412, + "precision": 0.9131843007748, + "recall": 0.9043965029247123, + "support": 15899.0 }, - "eval_accuracy": 0.9172705127809812, - "eval_loss": 0.604762613773346, + "eval_accuracy": 0.9121770203408466, + "eval_loss": 0.6214332580566406, "eval_macro avg": { - "f1-score": 0.8887961546485946, - "precision": 0.8827599985976573, - "recall": 0.8960152772980452, - "support": 32431.0 - }, - "eval_runtime": 4.8523, - "eval_samples_per_second": 16.487, - "eval_steps_per_second": 2.061, + "f1-score": 0.8798873769140577, + "precision": 0.8775981318956315, + "recall": 0.8822617587665385, + "support": 36380.0 + }, + "eval_runtime": 4.962, + "eval_samples_per_second": 16.324, + "eval_steps_per_second": 2.217, "eval_weighted avg": { - "f1-score": 0.918776399254959, - "precision": 0.92113920073012, - "recall": 0.9172705127809812, - "support": 32431.0 + "f1-score": 0.9126011797368843, + "precision": 0.9130808795913876, + "recall": 0.9121770203408466, + "support": 36380.0 }, "step": 1296 }, { "epoch": 17.0, "eval_Claim": { - "f1-score": 0.6726977499420088, - "precision": 0.7134071340713407, - "recall": 0.6363835856923414, - "support": 4557.0 + "f1-score": 0.6977198697068404, + "precision": 0.7524590163934426, + "recall": 0.6504048582995952, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.898166556218246, - "precision": 0.9003542958370239, - "recall": 0.8959894226531512, - "support": 2269.0 + "f1-score": 0.935002195871761, + "precision": 0.8998309382924767, + "recall": 0.9730347349177331, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, - "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "f1-score": 0.9994757732344791, + "precision": 0.9994757732344791, + "recall": 0.9994757732344791, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9172189909373732, - "precision": 0.9018486500864477, - "recall": 0.9331223338378973, - "support": 14534.0 + "f1-score": 0.9219572623103128, + "precision": 0.908120309926179, + "recall": 0.9362224039247752, + "support": 15899.0 }, - "eval_accuracy": 0.9116277635595572, - "eval_loss": 0.6666390299797058, + "eval_accuracy": 0.9228422210005498, + "eval_loss": 0.6075455546379089, "eval_macro avg": { - "f1-score": 0.8720095330048165, - "precision": 0.8789025199987031, - "recall": 0.8663512540265629, - "support": 32431.0 - }, - "eval_runtime": 4.8204, - "eval_samples_per_second": 16.596, - "eval_steps_per_second": 2.075, + "f1-score": 0.8885387752808483, + "precision": 0.8899715094616443, + "recall": 0.8897844425941457, + "support": 36380.0 + }, + "eval_runtime": 5.0046, + "eval_samples_per_second": 16.185, + "eval_steps_per_second": 2.198, "eval_weighted avg": { - "f1-score": 0.9097710324765249, - "precision": 0.9087714991080675, - "recall": 0.9116277635595572, - "support": 32431.0 + "f1-score": 0.9207454500932067, + "precision": 0.9200160099363349, + "recall": 0.9228422210005498, + "support": 36380.0 }, "step": 1377 }, { "epoch": 18.0, "eval_Claim": { - "f1-score": 0.7242054854157598, - "precision": 0.7184193478730296, - "recall": 0.7300855826201448, - "support": 4557.0 + "f1-score": 0.6748943740271293, + "precision": 0.7486433152442032, + "recall": 0.6143724696356275, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.91550204257149, - "precision": 0.8937867338371117, - "recall": 0.9382988100484795, - "support": 2269.0 + "f1-score": 0.9180620674257648, + "precision": 0.8974247053688346, + "recall": 0.9396709323583181, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, - "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "f1-score": 0.9997753145596164, + "precision": 0.9998501984870047, + "recall": 0.9997004418482738, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9218198185721211, - "precision": 0.9277948146083078, - "recall": 0.9159212880143113, - "support": 14534.0 + "f1-score": 0.9206027683147655, + "precision": 0.898945097099017, + "recall": 0.9433297691678721, + "support": 15899.0 }, - "eval_accuracy": 0.9200456353488946, - "eval_loss": 0.5898464918136597, + "eval_accuracy": 0.9191313908741067, + "eval_loss": 0.6335331201553345, "eval_macro avg": { - "f1-score": 0.8903705453702524, - "precision": 0.8850002240796123, - "recall": 0.8960538386514494, - "support": 32431.0 - }, - "eval_runtime": 4.8354, - "eval_samples_per_second": 16.545, - "eval_steps_per_second": 2.068, + "f1-score": 0.878333631081819, + "precision": 0.8862158290497649, + "recall": 0.874268403252523, + "support": 36380.0 + }, + "eval_runtime": 4.9489, + "eval_samples_per_second": 16.367, + "eval_steps_per_second": 2.223, "eval_weighted avg": { - "f1-score": 0.9202833015380038, - "precision": 0.9206441029524512, - "recall": 0.9200456353488946, - "support": 32431.0 + "f1-score": 0.9161451676737934, + "precision": 0.9154807870211, + "recall": 0.9191313908741067, + "support": 36380.0 }, "step": 1458 }, { "epoch": 18.52, - "grad_norm": 0.2594422399997711, + "grad_norm": 0.10011027753353119, "learning_rate": 1.2592592592592593e-05, - "loss": 0.0093, + "loss": 0.0091, "step": 1500 }, { "epoch": 19.0, "eval_Claim": { - "f1-score": 0.7198604752561586, - "precision": 0.7151830192765866, - "recall": 0.7245995172262454, - "support": 4557.0 + "f1-score": 0.7046035805626598, + "precision": 0.7439243924392439, + "recall": 0.6692307692307692, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9025769956002515, - "precision": 0.860223642172524, - "recall": 0.9493168796826796, - "support": 2269.0 + "f1-score": 0.9264422045893201, + "precision": 0.8727272727272727, + "recall": 0.9872029250457038, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9975098474215601, + "f1-score": 0.999213041034289, "precision": 1.0, - "recall": 0.9950320657573841, - "support": 11071.0 + "recall": 0.9984273197034375, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9212571111419453, - "precision": 0.9289911851126347, - "recall": 0.9136507499655979, - "support": 14534.0 + "f1-score": 0.921880854252529, + "precision": 0.9153078306156612, + "recall": 0.9285489653437323, + "support": 15899.0 }, - "eval_accuracy": 0.9173630168665783, - "eval_loss": 0.6108261346817017, + "eval_accuracy": 0.9225123694337548, + "eval_loss": 0.6328368782997131, "eval_macro avg": { - "f1-score": 0.8853011073549788, - "precision": 0.8760994616404363, - "recall": 0.8956498031579767, - "support": 32431.0 - }, - "eval_runtime": 4.8341, - "eval_samples_per_second": 16.549, - "eval_steps_per_second": 2.069, + "f1-score": 0.8880349201096995, + "precision": 0.8829898739455444, + "recall": 0.8958524948309107, + "support": 36380.0 + }, + "eval_runtime": 4.9991, + "eval_samples_per_second": 16.203, + "eval_steps_per_second": 2.2, "eval_weighted avg": { - "f1-score": 0.9176816552958715, - "precision": 0.9183773040411919, - "recall": 0.9173630168665783, - "support": 32431.0 + "f1-score": 0.9210355324440019, + "precision": 0.9205605544347316, + "recall": 0.9225123694337548, + "support": 36380.0 }, "step": 1539 }, { "epoch": 20.0, "eval_Claim": { - "f1-score": 0.7150392937883518, - "precision": 0.7018174133558749, - "recall": 0.7287689269256089, - "support": 4557.0 + "f1-score": 0.6881697835679765, + "precision": 0.7153778942769768, + "recall": 0.6629554655870445, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8972902097902098, - "precision": 0.8899003034243607, - "recall": 0.9048038783605112, - "support": 2269.0 + "f1-score": 0.9309440559440559, + "precision": 0.8919597989949749, + "recall": 0.973491773308958, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, - "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9997753313862053, + "precision": 0.9997753313862053, + "recall": 0.9997753313862053, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9203380437794403, - "precision": 0.9266285395452644, - "recall": 0.9141323792486583, - "support": 14534.0 + "f1-score": 0.9161451814768461, + "precision": 0.9115248116555632, + "recall": 0.92081262972514, + "support": 15899.0 }, - "eval_accuracy": 0.9162221331442139, - "eval_loss": 0.6273905038833618, + "eval_accuracy": 0.9179494227597581, + "eval_loss": 0.6887677907943726, "eval_macro avg": { - "f1-score": 0.8829747964440202, - "precision": 0.879586564081375, - "recall": 0.886542410305856, - "support": 32431.0 - }, - "eval_runtime": 4.8321, - "eval_samples_per_second": 16.556, - "eval_steps_per_second": 2.069, + "f1-score": 0.883758588093771, + "precision": 0.87965945907843, + "recall": 0.8892588000018369, + "support": 36380.0 + }, + "eval_runtime": 5.0229, + "eval_samples_per_second": 16.126, + "eval_steps_per_second": 2.19, "eval_weighted avg": { - "f1-score": 0.9168102169228345, - "precision": 0.9175167258142971, - "recall": 0.9162221331442139, - "support": 32431.0 + "f1-score": 0.916774507023963, + "precision": 0.9161052176591827, + "recall": 0.9179494227597581, + "support": 36380.0 }, "step": 1620 }, { "epoch": 21.0, "eval_Claim": { - "f1-score": 0.7209400804573365, - "precision": 0.6964614440580896, - "recall": 0.7472021066491112, - "support": 4557.0 + "f1-score": 0.6423871114136601, + "precision": 0.7307692307692307, + "recall": 0.573076923076923, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9076372817168105, - "precision": 0.8683574879227053, - "recall": 0.9506390480387836, - "support": 2269.0 + "f1-score": 0.8879392212725545, + "precision": 0.9239130434782609, + "recall": 0.8546617915904936, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.999186477447347, - "precision": 0.9999095431931253, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9984999624990625, + "precision": 1.0, + "recall": 0.9970044184827379, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9181764025650909, - "precision": 0.9355852317360566, - "recall": 0.9014036053392046, - "support": 14534.0 + "f1-score": 0.9172614007499698, + "precision": 0.883336245558856, + "recall": 0.9538964714761935, + "support": 15899.0 }, - "eval_accuracy": 0.916314637229811, - "eval_loss": 0.6812385320663452, + "eval_accuracy": 0.9120395821880154, + "eval_loss": 0.7241741418838501, "eval_macro avg": { - "f1-score": 0.8864850605466462, - "precision": 0.8750784267274943, - "recall": 0.8994273041789365, - "support": 32431.0 - }, - "eval_runtime": 4.8372, - "eval_samples_per_second": 16.539, - "eval_steps_per_second": 2.067, + "f1-score": 0.8615219239838117, + "precision": 0.8845046299515869, + "recall": 0.844659901156587, + "support": 36380.0 + }, + "eval_runtime": 4.9949, + "eval_samples_per_second": 16.217, + "eval_steps_per_second": 2.202, "eval_weighted avg": { - "f1-score": 0.917379120765938, - "precision": 0.9192399941849548, - "recall": 0.916314637229811, - "support": 32431.0 + "f1-score": 0.9079909938510604, + "precision": 0.9078802833224489, + "recall": 0.9120395821880154, + "support": 36380.0 }, "step": 1701 }, { "epoch": 22.0, "eval_Claim": { - "f1-score": 0.7066195738862937, - "precision": 0.7348341232227488, - "recall": 0.6804915514592934, - "support": 4557.0 + "f1-score": 0.6720596061734965, + "precision": 0.7086419753086419, + "recall": 0.6390688259109312, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9032682605834613, - "precision": 0.8991266375545851, - "recall": 0.9074482150727192, - "support": 2269.0 + "f1-score": 0.9051685393258427, + "precision": 0.8903625110521662, + "recall": 0.9204753199268738, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, - "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9997004867090977, + "precision": 0.9995507973347308, + "recall": 0.9998502209241369, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9243903268596307, - "precision": 0.914037801843008, - "recall": 0.9349800467868447, - "support": 14534.0 + "f1-score": 0.9166278528178853, + "precision": 0.9051882742548755, + "recall": 0.9283602742310837, + "support": 15899.0 }, - "eval_accuracy": 0.9189664210169283, - "eval_loss": 0.6876958012580872, + "eval_accuracy": 0.9148433205057724, + "eval_loss": 0.7163732051849365, "eval_macro avg": { - "f1-score": 0.8833774499368663, - "precision": 0.8869996406550854, - "recall": 0.8803460675018758, - "support": 32431.0 - }, - "eval_runtime": 4.8487, - "eval_samples_per_second": 16.499, - "eval_steps_per_second": 2.062, + "f1-score": 0.8733891212565805, + "precision": 0.8759358894876037, + "recall": 0.8719386602482564, + "support": 36380.0 + }, + "eval_runtime": 4.9476, + "eval_samples_per_second": 16.371, + "eval_steps_per_second": 2.223, "eval_weighted avg": { - "f1-score": 0.917861415342362, - "precision": 0.9171589791287257, - "recall": 0.9189664210169283, - "support": 32431.0 + "f1-score": 0.9132201772820274, + "precision": 0.9122428450081295, + "recall": 0.9148433205057724, + "support": 36380.0 }, "step": 1782 }, { "epoch": 23.0, "eval_Claim": { - "f1-score": 0.7040384395378104, - "precision": 0.7354206500956023, - "recall": 0.6752249286811499, - "support": 4557.0 + "f1-score": 0.7022826614861584, + "precision": 0.6750700280112045, + "recall": 0.7317813765182186, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.90536013400335, - "precision": 0.8623853211009175, - "recall": 0.9528426619656236, - "support": 2269.0 + "f1-score": 0.9080485115766262, + "precision": 0.8772901576480614, + "recall": 0.9410420475319927, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9991412429378531, - "precision": 0.9999095350099512, - "recall": 0.9983741306115076, - "support": 11071.0 + "f1-score": 0.9988753936122357, + "precision": 1.0, + "recall": 0.9977533138620535, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9212183436002738, - "precision": 0.9164510418085251, - "recall": 0.9260355029585798, - "support": 14534.0 + "f1-score": 0.9134190823574582, + "precision": 0.9295994789970694, + "recall": 0.8977923139820114, + "support": 15899.0 }, - "eval_accuracy": 0.9173630168665783, - "eval_loss": 0.6735763549804688, + "eval_accuracy": 0.9145409565695437, + "eval_loss": 0.695353627204895, "eval_macro avg": { - "f1-score": 0.8824395400198218, - "precision": 0.878541637003749, - "recall": 0.8881193060542152, - "support": 32431.0 - }, - "eval_runtime": 4.8285, - "eval_samples_per_second": 16.568, - "eval_steps_per_second": 2.071, + "f1-score": 0.8806564122581195, + "precision": 0.8704899161640838, + "recall": 0.892092262973569, + "support": 36380.0 + }, + "eval_runtime": 4.9626, + "eval_samples_per_second": 16.322, + "eval_steps_per_second": 2.217, "eval_weighted avg": { - "f1-score": 0.9161926989447983, - "precision": 0.9157214424409951, - "recall": 0.9173630168665783, - "support": 32431.0 + "f1-score": 0.9157921828580727, + "precision": 0.9177311412832249, + "recall": 0.9145409565695437, + "support": 36380.0 }, "step": 1863 }, { "epoch": 24.0, "eval_Claim": { - "f1-score": 0.7284623773173391, - "precision": 0.724040754389768, - "recall": 0.7329383366249725, - "support": 4557.0 + "f1-score": 0.6945771284068559, + "precision": 0.6463308349311487, + "recall": 0.7506072874493928, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8918801563178463, - "precision": 0.8789045785194695, - "recall": 0.9052446011458792, - "support": 2269.0 + "f1-score": 0.9287416071041802, + "precision": 0.8826677645121449, + "recall": 0.979890310786106, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, + "f1-score": 0.9992880428673137, "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "recall": 0.9985770987793006, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9255205276060909, - "precision": 0.9289526582103002, - "recall": 0.9221136645108022, - "support": 14534.0 + "f1-score": 0.9030182916923878, + "precision": 0.9339381720430108, + "recall": 0.8740801308258381, + "support": 15899.0 }, - "eval_accuracy": 0.920415651691283, - "eval_loss": 0.6322010159492493, + "eval_accuracy": 0.9093732820230896, + "eval_loss": 0.7689566016197205, "eval_macro avg": { - "f1-score": 0.8862736749148389, - "precision": 0.8829744977798845, - "recall": 0.8896902647425751, - "support": 32431.0 - }, - "eval_runtime": 4.8235, - "eval_samples_per_second": 16.586, - "eval_steps_per_second": 2.073, + "f1-score": 0.8814062675176844, + "precision": 0.8657341928715762, + "recall": 0.9007887069601594, + "support": 36380.0 + }, + "eval_runtime": 4.9669, + "eval_samples_per_second": 16.308, + "eval_steps_per_second": 2.215, "eval_weighted avg": { - "f1-score": 0.9206403732624274, - "precision": 0.9209116629411165, - "recall": 0.920415651691283, - "support": 32431.0 + "f1-score": 0.911596446033489, + "precision": 0.9160482240413489, + "recall": 0.9093732820230896, + "support": 36380.0 }, "step": 1944 }, { "epoch": 24.69, - "grad_norm": 0.6325204968452454, + "grad_norm": 0.003183299908414483, "learning_rate": 1.0123456790123458e-05, - "loss": 0.004, + "loss": 0.0045, "step": 2000 }, { "epoch": 25.0, "eval_Claim": { - "f1-score": 0.7145744441949248, - "precision": 0.7316624511381927, - "recall": 0.6982664033355278, - "support": 4557.0 + "f1-score": 0.6596491228070176, + "precision": 0.7481386392811297, + "recall": 0.5898785425101215, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9116772823779193, - "precision": 0.8795575583777141, - "recall": 0.9462318201851035, - "support": 2269.0 + "f1-score": 0.9256909832351607, + "precision": 0.9177897574123989, + "recall": 0.9337294332723949, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, - "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9998876614866129, + "precision": 0.9999251048532055, + "recall": 0.9998502209241369, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9222897565330862, - "precision": 0.920614245561116, - "recall": 0.9239713774597496, - "support": 14534.0 + "f1-score": 0.9174541242455649, + "precision": 0.8901046903649376, + "recall": 0.9465375180828983, + "support": 15899.0 }, - "eval_accuracy": 0.9192439332737196, - "eval_loss": 0.6634725332260132, + "eval_accuracy": 0.9169048927982408, + "eval_loss": 0.7267045378684998, "eval_macro avg": { - "f1-score": 0.8869432803810023, - "precision": 0.8829585637692556, - "recall": 0.8917335144172567, - "support": 32431.0 - }, - "eval_runtime": 4.8751, - "eval_samples_per_second": 16.41, - "eval_steps_per_second": 2.051, + "f1-score": 0.8756704729435891, + "precision": 0.888989547977918, + "recall": 0.867498928697388, + "support": 36380.0 + }, + "eval_runtime": 4.9577, + "eval_samples_per_second": 16.338, + "eval_steps_per_second": 2.219, "eval_weighted avg": { - "f1-score": 0.9186261381483827, - "precision": 0.9182914290271974, - "recall": 0.9192439332737196, - "support": 32431.0 + "f1-score": 0.9131990545133639, + "precision": 0.9128010792876335, + "recall": 0.9169048927982408, + "support": 36380.0 }, "step": 2025 }, { "epoch": 26.0, "eval_Claim": { - "f1-score": 0.714535519125683, - "precision": 0.7117352492924015, - "recall": 0.717357910906298, - "support": 4557.0 + "f1-score": 0.7098703888334996, + "precision": 0.6994106090373281, + "recall": 0.7206477732793523, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9060118543607113, - "precision": 0.8716904276985743, - "recall": 0.9431467606875276, - "support": 2269.0 + "f1-score": 0.9220324134910205, + "precision": 0.8851976450798991, + "recall": 0.9620658135283364, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9990506758283985, + "f1-score": 0.9995504607777028, "precision": 1.0, - "recall": 0.9981031523800922, - "support": 11071.0 + "recall": 0.9991013255448213, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9199431877230055, - "precision": 0.926393637061327, - "recall": 0.9135819457823036, - "support": 14534.0 + "f1-score": 0.9171909755322529, + "precision": 0.9268511977393873, + "recall": 0.9077300459148374, + "support": 15899.0 }, - "eval_accuracy": 0.9169313311337918, - "eval_loss": 0.6303670406341553, + "eval_accuracy": 0.9191313908741067, + "eval_loss": 0.6802815794944763, "eval_macro avg": { - "f1-score": 0.8848853092594496, - "precision": 0.8774548285130757, - "recall": 0.8930474424390553, - "support": 32431.0 - }, - "eval_runtime": 4.8348, - "eval_samples_per_second": 16.547, - "eval_steps_per_second": 2.068, + "f1-score": 0.887161059658619, + "precision": 0.8778648629641536, + "recall": 0.8973862395668368, + "support": 36380.0 + }, + "eval_runtime": 4.9696, + "eval_samples_per_second": 16.299, + "eval_steps_per_second": 2.213, "eval_weighted avg": { - "f1-score": 0.9171108994684886, - "precision": 0.9175310114557944, - "recall": 0.9169313311337918, - "support": 32431.0 + "f1-score": 0.9195597378864155, + "precision": 0.9203108314716255, + "recall": 0.9191313908741067, + "support": 36380.0 }, "step": 2106 }, { "epoch": 27.0, "eval_Claim": { - "f1-score": 0.7246061922868008, - "precision": 0.717512908777969, - "recall": 0.7318411235461927, - "support": 4557.0 + "f1-score": 0.698643105080467, + "precision": 0.7271731990365666, + "recall": 0.6722672064777327, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9032534246575342, - "precision": 0.8780690803162713, - "recall": 0.9299250771264874, - "support": 2269.0 + "f1-score": 0.9165368350767862, + "precision": 0.8932754880694144, + "recall": 0.9410420475319927, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9991864039052614, + "f1-score": 0.9991755359016639, "precision": 1.0, - "recall": 0.9983741306115076, - "support": 11071.0 + "recall": 0.9983524301655059, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9229063442015175, - "precision": 0.9295735324911008, - "recall": 0.9163341131140773, - "support": 14534.0 + "f1-score": 0.9206260132186058, + "precision": 0.9127155838536193, + "recall": 0.9286747594188314, + "support": 15899.0 }, - "eval_accuracy": 0.9193672720545157, - "eval_loss": 0.6543757319450378, + "eval_accuracy": 0.9201759208356239, + "eval_loss": 0.6692858934402466, "eval_macro avg": { - "f1-score": 0.8874880912627785, - "precision": 0.8812888803963352, - "recall": 0.8941186110995663, - "support": 32431.0 - }, - "eval_runtime": 4.8562, - "eval_samples_per_second": 16.474, - "eval_steps_per_second": 2.059, + "f1-score": 0.8837453723193808, + "precision": 0.8832910677399001, + "recall": 0.8850841108985157, + "support": 36380.0 + }, + "eval_runtime": 4.9586, + "eval_samples_per_second": 16.335, + "eval_steps_per_second": 2.218, "eval_weighted avg": { - "f1-score": 0.9197072530313251, - "precision": 0.9202142021449996, - "recall": 0.9193672720545157, - "support": 32431.0 + "f1-score": 0.9190682641369721, + "precision": 0.918388879571886, + "recall": 0.9201759208356239, + "support": 36380.0 }, "step": 2187 }, { "epoch": 28.0, "eval_Claim": { - "f1-score": 0.7182029107783169, - "precision": 0.6913705583756345, - "recall": 0.7472021066491112, - "support": 4557.0 + "f1-score": 0.6998441558441558, + "precision": 0.7188900747065101, + "recall": 0.6817813765182186, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9016903914590747, - "precision": 0.9101930848675348, - "recall": 0.8933450859409432, - "support": 2269.0 + "f1-score": 0.9112884834663626, + "precision": 0.9094219390077378, + "recall": 0.9131627056672761, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, - "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9997003745318352, + "precision": 0.9999250767962838, + "recall": 0.9994757732344791, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9179735039465906, - "precision": 0.9279437609841827, - "recall": 0.9082152194853447, - "support": 14534.0 + "f1-score": 0.9211232449297972, + "precision": 0.9139372175097518, + "recall": 0.9284231712686333, + "support": 15899.0 }, - "eval_accuracy": 0.9153587616786408, - "eval_loss": 0.67218017578125, + "eval_accuracy": 0.9200934579439253, + "eval_loss": 0.6845241189002991, "eval_macro avg": { - "f1-score": 0.8842746111505153, - "precision": 0.882376851056838, - "recall": 0.8868067171910112, - "support": 32431.0 - }, - "eval_runtime": 4.8266, - "eval_samples_per_second": 16.575, - "eval_steps_per_second": 2.072, + "f1-score": 0.8829890646930376, + "precision": 0.8855435770050709, + "recall": 0.8807107566721517, + "support": 36380.0 + }, + "eval_runtime": 4.9593, + "eval_samples_per_second": 16.333, + "eval_steps_per_second": 2.218, "eval_weighted avg": { - "f1-score": 0.9165029304654106, - "precision": 0.9180579805194509, - "recall": 0.9153587616786408, - "support": 32431.0 + "f1-score": 0.9193256433193232, + "precision": 0.9187416037176308, + "recall": 0.9200934579439253, + "support": 36380.0 }, "step": 2268 }, { "epoch": 29.0, "eval_Claim": { - "f1-score": 0.7078925956061839, - "precision": 0.7525951557093425, - "recall": 0.6682027649769585, - "support": 4557.0 + "f1-score": 0.6987782896054476, + "precision": 0.6914387633769322, + "recall": 0.7062753036437247, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9066271437645564, - "precision": 0.8724531377343113, - "recall": 0.9435874834728956, - "support": 2269.0 + "f1-score": 0.9134720700985761, + "precision": 0.8771560790912916, + "recall": 0.9529250457038391, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, - "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9993630811884155, + "precision": 0.9999250262408157, + "recall": 0.9988017673930952, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9239400224405835, - "precision": 0.9132889695503126, - "recall": 0.9348424384202559, - "support": 14534.0 + "f1-score": 0.9151595913446284, + "precision": 0.9233625712273513, + "recall": 0.907101075539342, + "support": 15899.0 }, - "eval_accuracy": 0.9197064537017051, - "eval_loss": 0.7057467103004456, + "eval_accuracy": 0.916245189664651, + "eval_loss": 0.7010424137115479, "eval_macro avg": { - "f1-score": 0.8844228500573508, - "precision": 0.8845843157484916, - "recall": 0.8862742858896889, - "support": 32431.0 - }, - "eval_runtime": 4.8302, - "eval_samples_per_second": 16.562, - "eval_steps_per_second": 2.07, + "f1-score": 0.8816932580592669, + "precision": 0.8729706099840976, + "recall": 0.8912757980700003, + "support": 36380.0 + }, + "eval_runtime": 4.9512, + "eval_samples_per_second": 16.36, + "eval_steps_per_second": 2.222, "eval_weighted avg": { - "f1-score": 0.9180734884048335, - "precision": 0.9174528746424986, - "recall": 0.9197064537017051, - "support": 32431.0 + "f1-score": 0.9165821661881187, + "precision": 0.9171925615027781, + "recall": 0.916245189664651, + "support": 36380.0 }, "step": 2349 }, { "epoch": 30.0, "eval_Claim": { - "f1-score": 0.7153448663221748, - "precision": 0.7327963176064441, - "recall": 0.6987052885670397, - "support": 4557.0 + "f1-score": 0.6948434432158727, + "precision": 0.7097318978256281, + "recall": 0.6805668016194332, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8877374784110534, - "precision": 0.8700804062632247, - "recall": 0.9061260467166152, - "support": 2269.0 + "f1-score": 0.9181940098345999, + "precision": 0.8985126859142607, + "recall": 0.9387568555758684, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, - "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9995504944560982, + "precision": 0.9999250543356066, + "recall": 0.9991762150827529, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9262746978050201, - "precision": 0.9220124071170496, - "recall": 0.9305765790560067, - "support": 14534.0 + "f1-score": 0.9180584714693073, + "precision": 0.91476208317722, + "recall": 0.9213787030630858, + "support": 15899.0 }, - "eval_accuracy": 0.9194597761401129, - "eval_loss": 0.6960250735282898, + "eval_accuracy": 0.9182792743265531, + "eval_loss": 0.7106152176856995, "eval_macro avg": { - "f1-score": 0.882147170239082, - "precision": 0.8812222827466796, - "recall": 0.8834680927570769, - "support": 32431.0 - }, - "eval_runtime": 4.8194, - "eval_samples_per_second": 16.6, - "eval_steps_per_second": 2.075, + "f1-score": 0.8826616047439695, + "precision": 0.8807329303131789, + "recall": 0.8849696438352851, + "support": 36380.0 + }, + "eval_runtime": 4.9623, + "eval_samples_per_second": 16.323, + "eval_steps_per_second": 2.217, "eval_weighted avg": { - "f1-score": 0.9188453276546991, - "precision": 0.918414282204774, - "recall": 0.9194597761401129, - "support": 32431.0 + "f1-score": 0.917667523182169, + "precision": 0.9172023898575309, + "recall": 0.9182792743265531, + "support": 36380.0 }, "step": 2430 }, { "epoch": 30.86, - "grad_norm": 0.009368549101054668, + "grad_norm": 0.0027945609763264656, "learning_rate": 7.654320987654322e-06, - "loss": 0.002, + "loss": 0.0031, "step": 2500 }, { "epoch": 31.0, "eval_Claim": { - "f1-score": 0.723667100130039, - "precision": 0.7148362235067437, - "recall": 0.7327188940092166, - "support": 4557.0 + "f1-score": 0.7131089189464049, + "precision": 0.7165338238299611, + "recall": 0.7097165991902834, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8953336240732664, - "precision": 0.8860595597755718, - "recall": 0.9048038783605112, - "support": 2269.0 + "f1-score": 0.9162036008001779, + "precision": 0.8918217221981826, + "recall": 0.9419561243144424, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9991864039052614, - "precision": 1.0, - "recall": 0.9983741306115076, - "support": 11071.0 + "f1-score": 0.9994380549207659, + "precision": 0.9999250374812594, + "recall": 0.9989515464689583, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9236620107868897, - "precision": 0.928283530229326, - "recall": 0.9190862804458511, - "support": 14534.0 + "f1-score": 0.9217583110130771, + "precision": 0.9235918161151806, + "recall": 0.9199320711994465, + "support": 15899.0 }, - "eval_accuracy": 0.9189664210169283, - "eval_loss": 0.6865902543067932, + "eval_accuracy": 0.9217152281473338, + "eval_loss": 0.6842674016952515, "eval_macro avg": { - "f1-score": 0.8854622847238642, - "precision": 0.8822948283779104, - "recall": 0.8887457958567716, - "support": 32431.0 - }, - "eval_runtime": 4.8664, - "eval_samples_per_second": 16.439, - "eval_steps_per_second": 2.055, + "f1-score": 0.8876272214201064, + "precision": 0.882968099906146, + "recall": 0.8926390852932827, + "support": 36380.0 + }, + "eval_runtime": 4.9684, + "eval_samples_per_second": 16.303, + "eval_steps_per_second": 2.214, "eval_weighted avg": { - "f1-score": 0.9193598504741339, - "precision": 0.9198190200735107, - "recall": 0.9189664210169283, - "support": 32431.0 + "f1-score": 0.9216037183150064, + "precision": 0.9215824169266671, + "recall": 0.9217152281473338, + "support": 36380.0 }, "step": 2511 }, { "epoch": 32.0, "eval_Claim": { - "f1-score": 0.7149137639499492, - "precision": 0.7350486787204451, - "recall": 0.695852534562212, - "support": 4557.0 + "f1-score": 0.6948140387637506, + "precision": 0.720086862106406, + "recall": 0.671255060728745, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8920269389528569, - "precision": 0.8796058269065982, - "recall": 0.9048038783605112, - "support": 2269.0 + "f1-score": 0.9145506734378449, + "precision": 0.8846646732165742, + "recall": 0.946526508226691, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.999322095177837, + "f1-score": 0.9991005172026085, "precision": 1.0, - "recall": 0.9986451088429229, - "support": 11071.0 + "recall": 0.9982026510896428, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9260107310071427, - "precision": 0.919942961906702, - "recall": 0.9321590752717765, - "support": 14534.0 + "f1-score": 0.9190726159230096, + "precision": 0.9131946600434647, + "recall": 0.9250267312409586, + "support": 15899.0 }, - "eval_accuracy": 0.9197372883969042, - "eval_loss": 0.6847501397132874, + "eval_accuracy": 0.918719076415613, + "eval_loss": 0.7216777205467224, "eval_macro avg": { - "f1-score": 0.8830683822719464, - "precision": 0.8836493668834363, - "recall": 0.8828651492593558, - "support": 32431.0 - }, - "eval_runtime": 4.8433, - "eval_samples_per_second": 16.518, - "eval_steps_per_second": 2.065, + "f1-score": 0.8818844613318033, + "precision": 0.8794865488416111, + "recall": 0.8852527378215094, + "support": 36380.0 + }, + "eval_runtime": 4.9621, + "eval_samples_per_second": 16.324, + "eval_steps_per_second": 2.217, "eval_weighted avg": { - "f1-score": 0.9189974415520827, - "precision": 0.9184697807200564, - "recall": 0.9197372883969042, - "support": 32431.0 + "f1-score": 0.9177224560676552, + "precision": 0.9171181227002353, + "recall": 0.918719076415613, + "support": 36380.0 }, "step": 2592 }, { "epoch": 33.0, "eval_Claim": { - "f1-score": 0.7068796869905445, - "precision": 0.7002583979328165, - "recall": 0.7136273864384464, - "support": 4557.0 + "f1-score": 0.7047761512982157, + "precision": 0.7463226974428604, + "recall": 0.6676113360323886, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8840610884061089, - "precision": 0.8634453781512605, - "recall": 0.9056853239312472, - "support": 2269.0 + "f1-score": 0.921081556386019, + "precision": 0.8873358746293943, + "recall": 0.9574954296160878, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9987338337704622, + "f1-score": 0.9991380279578758, "precision": 1.0, - "recall": 0.9974708698401229, - "support": 11071.0 + "recall": 0.9982775406275743, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9202020901100422, - "precision": 0.9256474519632414, - "recall": 0.9148204210816018, - "support": 14534.0 + "f1-score": 0.9233112623954739, + "precision": 0.9127842655193608, + "recall": 0.9340839046480911, + "support": 15899.0 }, - "eval_accuracy": 0.9141253738706793, - "eval_loss": 0.7124452590942383, + "eval_accuracy": 0.922869708631116, + "eval_loss": 0.7047613263130188, "eval_macro avg": { - "f1-score": 0.8774691748192894, - "precision": 0.8723378070118297, - "recall": 0.8829010003228546, - "support": 32431.0 - }, - "eval_runtime": 4.8097, - "eval_samples_per_second": 16.633, - "eval_steps_per_second": 2.079, + "f1-score": 0.8870767495093961, + "precision": 0.8866107093979039, + "recall": 0.8893670527310353, + "support": 36380.0 + }, + "eval_runtime": 4.9399, + "eval_samples_per_second": 16.397, + "eval_steps_per_second": 2.227, "eval_weighted avg": { - "f1-score": 0.9145072552354695, - "precision": 0.9150070965816288, - "recall": 0.9141253738706793, - "support": 32431.0 + "f1-score": 0.9213341528568705, + "precision": 0.9206619586736987, + "recall": 0.922869708631116, + "support": 36380.0 }, "step": 2673 }, { "epoch": 34.0, "eval_Claim": { - "f1-score": 0.709584487534626, - "precision": 0.7166517457475381, - "recall": 0.7026552556506473, - "support": 4557.0 + "f1-score": 0.704132582598534, + "precision": 0.7408897831433042, + "recall": 0.6708502024291498, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8878668418747262, - "precision": 0.8824553765781454, - "recall": 0.8933450859409432, - "support": 2269.0 + "f1-score": 0.9225707391971923, + "precision": 0.8869675242513707, + "recall": 0.9611517367458866, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9992316384180792, - "precision": 1.0, - "recall": 0.998464456688646, - "support": 11071.0 + "f1-score": 0.9989880439263896, + "precision": 0.9999249699879952, + "recall": 0.9980528720137797, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9230083030261442, - "precision": 0.9205447577333699, - "recall": 0.9254850694922251, - "support": 14534.0 + "f1-score": 0.9222287974584982, + "precision": 0.9134378084896347, + "recall": 0.9311906409208126, + "support": 15899.0 }, - "eval_accuracy": 0.9168388270481946, - "eval_loss": 0.6949135661125183, + "eval_accuracy": 0.9221825178669598, + "eval_loss": 0.7385492324829102, "eval_macro avg": { - "f1-score": 0.879922817713394, - "precision": 0.8799129700147633, - "recall": 0.8799874669431155, - "support": 32431.0 - }, - "eval_runtime": 4.8274, - "eval_samples_per_second": 16.572, - "eval_steps_per_second": 2.072, + "f1-score": 0.8869800407951536, + "precision": 0.8853050214680762, + "recall": 0.8903113630274071, + "support": 36380.0 + }, + "eval_runtime": 4.9565, + "eval_samples_per_second": 16.342, + "eval_steps_per_second": 2.219, "eval_weighted avg": { - "f1-score": 0.9165811266694699, - "precision": 0.9163538208419149, - "recall": 0.9168388270481946, - "support": 32431.0 + "f1-score": 0.9208082115651994, + "precision": 0.9201601515452535, + "recall": 0.9221825178669598, + "support": 36380.0 }, "step": 2754 }, { "epoch": 35.0, "eval_Claim": { - "f1-score": 0.7161833075049812, - "precision": 0.7225820862184499, - "recall": 0.7098968619705947, - "support": 4557.0 + "f1-score": 0.7115286429081996, + "precision": 0.7179064496187925, + "recall": 0.7052631578947368, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9034467994005567, - "precision": 0.8784346378018318, - "recall": 0.9299250771264874, - "support": 2269.0 + "f1-score": 0.9221824686940967, + "precision": 0.9028021015761821, + "recall": 0.9424131627056673, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9995933673699905, - "precision": 1.0, - "recall": 0.9991870653057537, - "support": 11071.0 + "f1-score": 0.9993630811884155, + "precision": 0.9999250262408157, + "recall": 0.9988017673930952, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9223401323042998, - "precision": 0.9237405106970324, - "recall": 0.9209439933947984, - "support": 14534.0 + "f1-score": 0.9208904540309396, + "precision": 0.9207167557371896, + "recall": 0.9210642178753381, + "support": 15899.0 }, - "eval_accuracy": 0.9186272393697388, - "eval_loss": 0.6991766691207886, + "eval_accuracy": 0.9215777899945025, + "eval_loss": 0.7338384389877319, "eval_macro avg": { - "f1-score": 0.885390901644957, - "precision": 0.8811893086793285, - "recall": 0.8899882494494086, - "support": 32431.0 - }, - "eval_runtime": 4.8369, - "eval_samples_per_second": 16.539, - "eval_steps_per_second": 2.067, + "f1-score": 0.8884911617054129, + "precision": 0.8853375832932451, + "recall": 0.8918855764672093, + "support": 36380.0 + }, + "eval_runtime": 4.9467, + "eval_samples_per_second": 16.375, + "eval_steps_per_second": 2.224, "eval_weighted avg": { - "f1-score": 0.9184224283310388, - "precision": 0.9183379896562086, - "recall": 0.9186272393697388, - "support": 32431.0 + "f1-score": 0.9213419265864765, + "precision": 0.9211727167186562, + "recall": 0.9215777899945025, + "support": 36380.0 }, "step": 2835 }, { "epoch": 36.0, "eval_Claim": { - "f1-score": 0.7000227946204696, - "precision": 0.7282428266540194, - "recall": 0.673908272986614, - "support": 4557.0 + "f1-score": 0.7146180690189995, + "precision": 0.6856398809523809, + "recall": 0.7461538461538462, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9057333047025982, - "precision": 0.8831658291457286, - "recall": 0.9294843543411194, - "support": 2269.0 + "f1-score": 0.9238578680203046, + "precision": 0.893299189073837, + "recall": 0.956581352833638, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9992505433560669, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.998502209241369, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9210652099692729, - "precision": 0.9141366223908919, - "recall": 0.9280996284574102, - "support": 14534.0 + "f1-score": 0.9148493291062222, + "precision": 0.9318893528183716, + "recall": 0.8984212843575068, + "support": 15899.0 }, - "eval_accuracy": 0.9169930005241899, - "eval_loss": 0.7325878739356995, + "eval_accuracy": 0.9179769103903244, + "eval_loss": 0.7659569978713989, "eval_macro avg": { - "f1-score": 0.8816940360534948, - "precision": 0.8813863195476599, - "recall": 0.8828504824270013, - "support": 32431.0 - }, - "eval_runtime": 4.8521, - "eval_samples_per_second": 16.488, - "eval_steps_per_second": 2.061, + "f1-score": 0.8881439523753982, + "precision": 0.8777071057111474, + "recall": 0.89991467314659, + "support": 36380.0 + }, + "eval_runtime": 4.9595, + "eval_samples_per_second": 16.332, + "eval_steps_per_second": 2.218, "eval_weighted avg": { - "f1-score": 0.9158636638576222, - "precision": 0.9151604174161526, - "recall": 0.9169930005241899, - "support": 32431.0 + "f1-score": 0.9191807659449058, + "precision": 0.9211299741082354, + "recall": 0.9179769103903244, + "support": 36380.0 }, "step": 2916 }, { "epoch": 37.0, "eval_Claim": { - "f1-score": 0.7043226381461676, - "precision": 0.7153202081918986, - "recall": 0.6936581084046521, - "support": 4557.0 + "f1-score": 0.7063829787234043, + "precision": 0.7443946188340808, + "recall": 0.6720647773279352, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8989154013015185, - "precision": 0.8850918410935498, - "recall": 0.9131776112825033, - "support": 2269.0 + "f1-score": 0.9240397065170479, + "precision": 0.875306623058054, + "recall": 0.9785191956124314, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9987627938364638, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.9975286452482588, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9208855328642527, - "precision": 0.9187726868022739, - "recall": 0.9230081188936288, - "support": 14534.0 + "f1-score": 0.9224097588369263, + "precision": 0.9151293797201931, + "recall": 0.9298069060947229, + "support": 15899.0 }, - "eval_accuracy": 0.91634547192501, - "eval_loss": 0.7374063730239868, + "eval_accuracy": 0.9225948323254536, + "eval_loss": 0.760911226272583, "eval_macro avg": { - "f1-score": 0.8810196018083942, - "precision": 0.8797961840219305, - "recall": 0.8824383781259115, - "support": 32431.0 - }, - "eval_runtime": 4.8356, - "eval_samples_per_second": 16.544, - "eval_steps_per_second": 2.068, + "f1-score": 0.8878988094784606, + "precision": 0.883707655403082, + "recall": 0.8944798810708371, + "support": 36380.0 + }, + "eval_runtime": 4.9543, + "eval_samples_per_second": 16.349, + "eval_steps_per_second": 2.22, "eval_weighted avg": { - "f1-score": 0.9159103209784385, - "precision": 0.9155570227916499, - "recall": 0.91634547192501, - "support": 32431.0 + "f1-score": 0.9211985468554017, + "precision": 0.9207015480061224, + "recall": 0.9225948323254536, + "support": 36380.0 }, "step": 2997 }, { "epoch": 37.04, - "grad_norm": 0.13183258473873138, + "grad_norm": 0.003850826993584633, "learning_rate": 5.185185185185185e-06, - "loss": 0.0012, + "loss": 0.0007, "step": 3000 }, { "epoch": 38.0, "eval_Claim": { - "f1-score": 0.6968245125348189, - "precision": 0.7077863286555002, - "recall": 0.6861970594689488, - "support": 4557.0 + "f1-score": 0.7178156146179401, + "precision": 0.7367860187553282, + "recall": 0.6997975708502024, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8926233085988651, - "precision": 0.8841331603977518, - "recall": 0.9012780960775673, - "support": 2269.0 + "f1-score": 0.9259421560035057, + "precision": 0.8893097643097643, + "recall": 0.9657221206581352, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, - "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "f1-score": 0.9997003745318352, + "precision": 0.9999250767962838, + "recall": 0.9994757732344791, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9194212042243862, - "precision": 0.9164046479835953, - "recall": 0.922457685427274, - "support": 14534.0 + "f1-score": 0.9237383881496359, + "precision": 0.9218290009395553, + "recall": 0.9256557016164538, + "support": 15899.0 }, - "eval_accuracy": 0.9142178779562764, - "eval_loss": 0.7381414175033569, + "eval_accuracy": 0.9244914788345244, + "eval_loss": 0.7453881502151489, "eval_macro avg": { - "f1-score": 0.8772059650699272, - "precision": 0.8770810342592118, - "recall": 0.877460628724163, - "support": 32431.0 - }, - "eval_runtime": 4.8281, - "eval_samples_per_second": 16.57, - "eval_steps_per_second": 2.071, + "f1-score": 0.8917991333257291, + "precision": 0.8869624652002329, + "recall": 0.8976627915898177, + "support": 36380.0 + }, + "eval_runtime": 4.9642, + "eval_samples_per_second": 16.317, + "eval_steps_per_second": 2.216, "eval_weighted avg": { - "f1-score": 0.9137602710507441, - "precision": 0.9133700963405134, - "recall": 0.9142178779562764, - "support": 32431.0 + "f1-score": 0.9237901678906805, + "precision": 0.9234109822253945, + "recall": 0.9244914788345244, + "support": 36380.0 }, "step": 3078 }, { "epoch": 39.0, "eval_Claim": { - "f1-score": 0.6997294860234445, - "precision": 0.719351100811124, - "recall": 0.6811498793065613, - "support": 4557.0 + "f1-score": 0.704778156996587, + "precision": 0.7448151487826871, + "recall": 0.668825910931174, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9017094017094017, - "precision": 0.8751555371215264, - "recall": 0.9299250771264874, - "support": 2269.0 + "f1-score": 0.9222684475585724, + "precision": 0.8852459016393442, + "recall": 0.9625228519195612, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9992880428673137, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.9985770987793006, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9202920909184409, - "precision": 0.9171165015374103, - "recall": 0.9234897481766892, - "support": 14534.0 + "f1-score": 0.9231870525988172, + "precision": 0.9137453021994948, + "recall": 0.9328259638971005, + "support": 15899.0 }, - "eval_accuracy": 0.9159754555826216, - "eval_loss": 0.7733151316642761, + "eval_accuracy": 0.9228971962616822, + "eval_loss": 0.7312874794006348, "eval_macro avg": { - "f1-score": 0.8804214533932313, - "precision": 0.8779057848675151, - "recall": 0.8836185946331498, - "support": 32431.0 - }, - "eval_runtime": 4.9105, - "eval_samples_per_second": 16.292, - "eval_steps_per_second": 2.036, + "f1-score": 0.8873804250053225, + "precision": 0.8859515881553816, + "recall": 0.8906879563817841, + "support": 36380.0 + }, + "eval_runtime": 4.966, + "eval_samples_per_second": 16.311, + "eval_steps_per_second": 2.215, "eval_weighted avg": { - "f1-score": 0.9151944475074327, - "precision": 0.9146860137975935, - "recall": 0.9159754555826216, - "support": 32431.0 + "f1-score": 0.9214065872594044, + "precision": 0.9207515510567105, + "recall": 0.9228971962616822, + "support": 36380.0 }, "step": 3159 }, { "epoch": 40.0, "eval_Claim": { - "f1-score": 0.6766934825696631, - "precision": 0.7218905472636816, - "recall": 0.6368224709238535, - "support": 4557.0 + "f1-score": 0.7213450891941768, + "precision": 0.7307852098047362, + "recall": 0.7121457489878542, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8926333981421473, - "precision": 0.8754237288135593, - "recall": 0.9105332745702953, - "support": 2269.0 + "f1-score": 0.9265486725663717, + "precision": 0.8979416809605489, + "recall": 0.9570383912248629, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999096657633243, + "f1-score": 0.9992880428673137, "precision": 1.0, - "recall": 0.999819347845723, - "support": 11071.0 + "recall": 0.9985770987793006, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9188236888467272, - "precision": 0.9050861033239888, - "recall": 0.9329847254713086, - "support": 14534.0 + "f1-score": 0.9231736847070662, + "precision": 0.9231446540880504, + "recall": 0.9232027171520222, + "support": 15899.0 }, - "eval_accuracy": 0.9126144738059264, - "eval_loss": 0.8124102354049683, + "eval_accuracy": 0.9242440901594282, + "eval_loss": 0.7378404140472412, "eval_macro avg": { - "f1-score": 0.8720150588304654, - "precision": 0.8756000948503074, - "recall": 0.870039954702795, - "support": 32431.0 - }, - "eval_runtime": 4.8273, - "eval_samples_per_second": 16.573, - "eval_steps_per_second": 2.072, + "f1-score": 0.8925888723337321, + "precision": 0.8879678862133338, + "recall": 0.89774098903601, + "support": 36380.0 + }, + "eval_runtime": 4.9789, + "eval_samples_per_second": 16.269, + "eval_steps_per_second": 2.209, "eval_weighted avg": { - "f1-score": 0.9106490945027469, - "precision": 0.9096701640488858, - "recall": 0.9126144738059264, - "support": 32431.0 + "f1-score": 0.9239077758592451, + "precision": 0.9237177622243814, + "recall": 0.9242440901594282, + "support": 36380.0 }, "step": 3240 }, { "epoch": 41.0, "eval_Claim": { - "f1-score": 0.6830284788145403, - "precision": 0.7228620436167607, - "recall": 0.6473557164801405, - "support": 4557.0 + "f1-score": 0.7108195355618034, + "precision": 0.731932232468368, + "recall": 0.6908906882591093, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9045826876245294, - "precision": 0.9088078291814946, - "recall": 0.9003966505068312, - "support": 2269.0 + "f1-score": 0.9284467713787086, + "precision": 0.8881469115191987, + "recall": 0.9725776965265083, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.999213041034289, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.9984273197034375, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9190962592166678, - "precision": 0.9038717402873869, - "recall": 0.9348424384202559, - "support": 14534.0 + "f1-score": 0.9213497240341193, + "precision": 0.9187566451935706, + "recall": 0.9239574816026165, + "support": 15899.0 }, - "eval_accuracy": 0.9142487126514754, - "eval_loss": 0.8041703104972839, + "eval_accuracy": 0.9225673446948873, + "eval_loss": 0.7535675168037415, "eval_macro avg": { - "f1-score": 0.876665565144344, - "precision": 0.8838854032714106, - "recall": 0.8706261198325222, - "support": 32431.0 - }, - "eval_runtime": 4.8224, - "eval_samples_per_second": 16.589, - "eval_steps_per_second": 2.074, + "f1-score": 0.8899572680022301, + "precision": 0.8847089472952844, + "recall": 0.8964632965229179, + "support": 36380.0 + }, + "eval_runtime": 4.9535, + "eval_samples_per_second": 16.352, + "eval_steps_per_second": 2.221, "eval_weighted avg": { - "f1-score": 0.9125128397228092, - "precision": 0.911598136675134, - "recall": 0.9142487126514754, - "support": 32431.0 + "f1-score": 0.921768033007181, + "precision": 0.9213667557100144, + "recall": 0.9225673446948873, + "support": 36380.0 }, "step": 3321 }, { "epoch": 42.0, "eval_Claim": { - "f1-score": 0.6939400568803326, - "precision": 0.6918211559432933, - "recall": 0.696071977177968, - "support": 4557.0 + "f1-score": 0.7140475197999167, + "precision": 0.7358247422680413, + "recall": 0.6935222672064777, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8969750109601051, - "precision": 0.8922808547754034, - "recall": 0.9017188188629353, - "support": 2269.0 + "f1-score": 0.9298360655737705, + "precision": 0.8910766652702137, + "recall": 0.9721206581352834, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9992880428673137, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.9985770987793006, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9167729262156668, - "precision": 0.9183870745011393, - "recall": 0.9151644419980735, - "support": 14534.0 + "f1-score": 0.9222619271519027, + "precision": 0.919265137786665, + "recall": 0.9252783193911567, + "support": 15899.0 }, - "eval_accuracy": 0.9123677962443342, - "eval_loss": 0.7728341221809387, + "eval_accuracy": 0.9235294117647059, + "eval_loss": 0.7416272759437561, "eval_macro avg": { - "f1-score": 0.8769107072444358, - "precision": 0.8756222713049591, - "recall": 0.8782162279904596, - "support": 32431.0 - }, - "eval_runtime": 4.8152, - "eval_samples_per_second": 16.614, - "eval_steps_per_second": 2.077, + "f1-score": 0.8913583888482258, + "precision": 0.88654163633123, + "recall": 0.8973745858780546, + "support": 36380.0 + }, + "eval_runtime": 4.9829, + "eval_samples_per_second": 16.256, + "eval_steps_per_second": 2.208, "eval_weighted avg": { - "f1-score": 0.9124725980114123, - "precision": 0.9125852427590434, - "recall": 0.9123677962443342, - "support": 32431.0 + "f1-score": 0.9227160988312902, + "precision": 0.9222937437076839, + "recall": 0.9235294117647059, + "support": 36380.0 }, "step": 3402 }, { "epoch": 43.0, "eval_Claim": { - "f1-score": 0.7087594486438418, - "precision": 0.7181797702185176, - "recall": 0.6995830590300637, - "support": 4557.0 + "f1-score": 0.7126268438121143, + "precision": 0.7373890452478892, + "recall": 0.6894736842105263, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9011231006386258, - "precision": 0.9005281690140845, - "recall": 0.9017188188629353, - "support": 2269.0 + "f1-score": 0.9299279633267845, + "precision": 0.8900961136648559, + "recall": 0.973491773308958, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9991380279578758, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.9982775406275743, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9225603070175438, - "precision": 0.9189078498293516, - "recall": 0.9262419155084629, - "support": 14534.0 + "f1-score": 0.9221279393806557, + "precision": 0.9181319366504551, + "recall": 0.9261588779168501, + "support": 15899.0 }, - "eval_accuracy": 0.9178255372945638, - "eval_loss": 0.7432543039321899, + "eval_accuracy": 0.9233369983507421, + "eval_loss": 0.7578195333480835, "eval_macro avg": { - "f1-score": 0.8830994228054124, - "precision": 0.8844039472654884, - "recall": 0.8818633668310809, - "support": 32431.0 - }, - "eval_runtime": 4.8581, - "eval_samples_per_second": 16.467, - "eval_steps_per_second": 2.058, + "f1-score": 0.8909551936193576, + "precision": 0.8864042738908001, + "recall": 0.8968504690159773, + "support": 36380.0 + }, + "eval_runtime": 4.9557, + "eval_samples_per_second": 16.345, + "eval_steps_per_second": 2.22, "eval_weighted avg": { - "f1-score": 0.9174387654537162, - "precision": 0.9170993900218476, - "recall": 0.9178255372945638, - "support": 32431.0 + "f1-score": 0.9224150958693075, + "precision": 0.9219519472520303, + "recall": 0.9233369983507421, + "support": 36380.0 }, "step": 3483 }, { "epoch": 43.21, - "grad_norm": 0.0030191524419933558, + "grad_norm": 0.03563033416867256, "learning_rate": 2.7160493827160496e-06, - "loss": 0.0009, + "loss": 0.0003, "step": 3500 }, { "epoch": 44.0, "eval_Claim": { - "f1-score": 0.7071469411092054, - "precision": 0.7382999044890163, - "recall": 0.6785165679174896, - "support": 4557.0 + "f1-score": 0.7031551835157759, + "precision": 0.7482868889904066, + "recall": 0.6631578947368421, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.9027962716378164, - "precision": 0.9092534644613322, - "recall": 0.8964301454385192, - "support": 2269.0 + "f1-score": 0.924901185770751, + "precision": 0.8901098901098901, + "recall": 0.9625228519195612, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9992505433560669, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.998502209241369, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9254835425856802, - "precision": 0.9130289234065345, - "recall": 0.9382826475849732, - "support": 14534.0 + "f1-score": 0.9228619340413639, + "precision": 0.9114273446604919, + "recall": 0.9345870809484873, + "support": 15899.0 }, - "eval_accuracy": 0.9198914618728994, - "eval_loss": 0.7702658772468567, + "eval_accuracy": 0.922869708631116, + "eval_loss": 0.7405058741569519, "eval_macro avg": { - "f1-score": 0.8838453975635852, - "precision": 0.8901455730892207, - "recall": 0.8782847587159609, - "support": 32431.0 - }, - "eval_runtime": 4.8623, - "eval_samples_per_second": 16.453, - "eval_steps_per_second": 2.057, + "f1-score": 0.8875422116709895, + "precision": 0.8874560309401971, + "recall": 0.8896925092115648, + "support": 36380.0 + }, + "eval_runtime": 4.9702, + "eval_samples_per_second": 16.297, + "eval_steps_per_second": 2.213, "eval_weighted avg": { - "f1-score": 0.9186392999395205, - "precision": 0.9179023510964751, - "recall": 0.9198914618728994, - "support": 32431.0 + "f1-score": 0.9211886969706305, + "precision": 0.9205025020321663, + "recall": 0.922869708631116, + "support": 36380.0 }, "step": 3564 }, { "epoch": 45.0, "eval_Claim": { - "f1-score": 0.7032942490228922, - "precision": 0.7160072760345612, - "recall": 0.6910247970155804, - "support": 4557.0 + "f1-score": 0.7141807494489346, + "precision": 0.7416612164813604, + "recall": 0.6886639676113361, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8986205386468141, - "precision": 0.8929503916449086, - "recall": 0.9043631555751432, - "support": 2269.0 + "f1-score": 0.9250714128762909, + "precision": 0.8908167583580194, + "recall": 0.9620658135283364, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9999548349216385, + "f1-score": 0.9991005172026085, "precision": 1.0, - "recall": 0.9999096739228616, - "support": 11071.0 + "recall": 0.9982026510896428, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9215384088496181, - "precision": 0.9174224343675418, - "recall": 0.9256914820421082, - "support": 14534.0 + "f1-score": 0.9235625000000001, + "precision": 0.9177690826656729, + "recall": 0.9294295238694258, + "support": 15899.0 }, - "eval_accuracy": 0.9165613147914032, - "eval_loss": 0.7518438696861267, + "eval_accuracy": 0.9239417262231996, + "eval_loss": 0.7358143329620361, "eval_macro avg": { - "f1-score": 0.8808520078602406, - "precision": 0.8815950255117528, - "recall": 0.8802472771389234, - "support": 32431.0 - }, - "eval_runtime": 4.8413, - "eval_samples_per_second": 16.524, - "eval_steps_per_second": 2.066, + "f1-score": 0.8904787948819586, + "precision": 0.8875617643762631, + "recall": 0.8945904890246853, + "support": 36380.0 + }, + "eval_runtime": 4.9567, + "eval_samples_per_second": 16.341, + "eval_steps_per_second": 2.219, "eval_weighted avg": { - "f1-score": 0.9160377757893605, - "precision": 0.9155982626693486, - "recall": 0.9165613147914032, - "support": 32431.0 + "f1-score": 0.9229471838196122, + "precision": 0.9224168257835843, + "recall": 0.9239417262231996, + "support": 36380.0 }, "step": 3645 }, { "epoch": 46.0, "eval_Claim": { - "f1-score": 0.7006939780613387, - "precision": 0.7151016678090016, - "recall": 0.6868553873162168, - "support": 4557.0 + "f1-score": 0.7153789705419855, + "precision": 0.7419004131332898, + "recall": 0.6906882591093118, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8954833078769365, - "precision": 0.8867761452031115, - "recall": 0.9043631555751432, - "support": 2269.0 + "f1-score": 0.9296449575255936, + "precision": 0.8880565959217644, + "recall": 0.9753199268738574, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9997289483194796, + "f1-score": 0.9989879680647701, "precision": 1.0, - "recall": 0.9994580435371692, - "support": 11071.0 + "recall": 0.9979779824758481, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9213598548392617, - "precision": 0.9169335604770017, - "recall": 0.9258290904086969, - "support": 14534.0 + "f1-score": 0.9230384025539106, + "precision": 0.9186394218788936, + "recall": 0.9274797157053902, + "support": 15899.0 }, - "eval_accuracy": 0.9158829514970245, - "eval_loss": 0.750240683555603, + "eval_accuracy": 0.9240791643760308, + "eval_loss": 0.7376746535301208, "eval_macro avg": { - "f1-score": 0.8793165222742542, - "precision": 0.8797028433722787, - "recall": 0.8791264192093065, - "support": 32431.0 - }, - "eval_runtime": 4.8402, - "eval_samples_per_second": 16.528, - "eval_steps_per_second": 2.066, + "f1-score": 0.891762574671565, + "precision": 0.887149107733487, + "recall": 0.8978664710411017, + "support": 36380.0 + }, + "eval_runtime": 4.9655, + "eval_samples_per_second": 16.313, + "eval_steps_per_second": 2.215, "eval_weighted avg": { - "f1-score": 0.9152957787511048, - "precision": 0.9148199482484113, - "recall": 0.9158829514970245, - "support": 32431.0 + "f1-score": 0.9231146009158029, + "precision": 0.9226636624850962, + "recall": 0.9240791643760308, + "support": 36380.0 }, "step": 3726 }, { "epoch": 47.0, "eval_Claim": { - "f1-score": 0.7070640176600441, - "precision": 0.7113035753941817, - "recall": 0.7028746982664034, - "support": 4557.0 + "f1-score": 0.7173056994818653, + "precision": 0.7348195329087048, + "recall": 0.7006072874493927, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8924216027874564, - "precision": 0.8820490744726647, - "recall": 0.9030409872190392, - "support": 2269.0 + "f1-score": 0.9317934190455436, + "precision": 0.8904623073719283, + "recall": 0.9771480804387569, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9995029596493606, + "f1-score": 0.999213041034289, "precision": 1.0, - "recall": 0.9990064131514769, - "support": 11071.0 + "recall": 0.9984273197034375, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.921558513016266, - "precision": 0.9212100378136817, - "recall": 0.9219072519609193, - "support": 14534.0 + "f1-score": 0.9222892323156174, + "precision": 0.9211896843822551, + "recall": 0.9233914082646707, + "support": 15899.0 }, - "eval_accuracy": 0.9161296290586167, - "eval_loss": 0.7462829351425171, + "eval_accuracy": 0.9239142385926333, + "eval_loss": 0.7394932508468628, "eval_macro avg": { - "f1-score": 0.8801367732782818, - "precision": 0.8786406719201321, - "recall": 0.8817073376494596, - "support": 32431.0 - }, - "eval_runtime": 4.8323, - "eval_samples_per_second": 16.555, - "eval_steps_per_second": 2.069, + "f1-score": 0.8926503479693289, + "precision": 0.8866178811657222, + "recall": 0.8998935239640644, + "support": 36380.0 + }, + "eval_runtime": 4.9785, + "eval_samples_per_second": 16.27, + "eval_steps_per_second": 2.21, "eval_weighted avg": { - "f1-score": 0.9159885307162297, - "precision": 0.9158720493550557, - "recall": 0.9161296290586167, - "support": 32431.0 + "f1-score": 0.9232606486484036, + "precision": 0.9229613747414034, + "recall": 0.9239142385926333, + "support": 36380.0 }, "step": 3807 }, { "epoch": 48.0, "eval_Claim": { - "f1-score": 0.702890432444544, - "precision": 0.7180132753490501, - "recall": 0.6883914856265086, - "support": 4557.0 + "f1-score": 0.7136500051743766, + "precision": 0.730044463264874, + "recall": 0.6979757085020243, + "support": 4940.0 }, "eval_MajorClaim": { - "f1-score": 0.8967924939995635, - "precision": 0.8880726015557476, - "recall": 0.9056853239312472, - "support": 2269.0 + "f1-score": 0.9288503253796095, + "precision": 0.8839801816680429, + "recall": 0.9785191956124314, + "support": 2188.0 }, "eval_O": { - "f1-score": 0.9998644925245043, + "f1-score": 0.9992505433560669, "precision": 1.0, - "recall": 0.9997290217685846, - "support": 11071.0 + "recall": 0.998502209241369, + "support": 13353.0 }, "eval_Premise": { - "f1-score": 0.9222290682549462, - "precision": 0.9176430517711172, - "recall": 0.926861153158112, - "support": 14534.0 + "f1-score": 0.9214804565894154, + "precision": 0.9213935354043516, + "recall": 0.9215673941757343, + "support": 15899.0 }, - "eval_accuracy": 0.9167463229625975, - "eval_loss": 0.7590324282646179, + "eval_accuracy": 0.922869708631116, + "eval_loss": 0.7407149076461792, "eval_macro avg": { - "f1-score": 0.8804441218058895, - "precision": 0.8809322321689788, - "recall": 0.8801667461211131, - "support": 32431.0 - }, - "eval_runtime": 4.831, - "eval_samples_per_second": 16.56, - "eval_steps_per_second": 2.07, + "f1-score": 0.8908078326248671, + "precision": 0.8838545450843172, + "recall": 0.8991411268828897, + "support": 36380.0 + }, + "eval_runtime": 4.9928, + "eval_samples_per_second": 16.223, + "eval_steps_per_second": 2.203, "eval_weighted avg": { - "f1-score": 0.916131816604205, - "precision": 0.9156377337466323, - "recall": 0.9167463229625975, - "support": 32431.0 + "f1-score": 0.9222475487146972, + "precision": 0.9220122074054958, + "recall": 0.922869708631116, + "support": 36380.0 }, "step": 3888 } @@ -2223,7 +2223,7 @@ "num_input_tokens_seen": 0, "num_train_epochs": 50, "save_steps": 500, - "total_flos": 6901859170886400.0, + "total_flos": 6880424825635200.0, "train_batch_size": 4, "trial_name": null, "trial_params": null