{ "best_metric": null, "best_model_checkpoint": null, "epoch": 12.0, "eval_steps": 500, "global_step": 492, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 1.0, "eval_Claim": { "f1-score": 0.31881702957426067, "precision": 0.49395770392749244, "recall": 0.23536468330134358, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.5882107894188711, "precision": 0.5330313325783315, "recall": 0.6561338289962825, "support": 2152.0 }, "eval_O": { "f1-score": 0.8686845568461407, "precision": 0.9200096957944491, "recall": 0.82278343810969, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8547340147728121, "precision": 0.777574153261386, "recall": 0.9488942267870455, "support": 12073.0 }, "eval_accuracy": 0.7762772004779318, "eval_loss": 0.5690305829048157, "eval_macro avg": { "f1-score": 0.6576115976530211, "precision": 0.6811432213904147, "recall": 0.6657940442985903, "support": 27619.0 }, "eval_runtime": 1.3776, "eval_samples_per_second": 58.074, "eval_steps_per_second": 7.259, "eval_weighted avg": { "f1-score": 0.7577517824653167, "precision": 0.7632992267425562, "recall": 0.7762772004779318, "support": 27619.0 }, "step": 41 }, { "epoch": 2.0, "eval_Claim": { "f1-score": 0.5093288777499304, "precision": 0.6068347710683477, "recall": 0.43881957773512476, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7403386886669561, "precision": 0.6947840260798696, "recall": 0.7922862453531598, "support": 2152.0 }, "eval_O": { "f1-score": 0.8998096943915818, "precision": 0.930324074074074, "recall": 0.8712334706264904, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8735146966854284, "precision": 0.8270298275479239, "recall": 0.9255363207156465, "support": 12073.0 }, "eval_accuracy": 0.823563488902567, "eval_loss": 0.44297581911087036, "eval_macro avg": { "f1-score": 0.7557479893734742, "precision": 0.7647431746925538, "recall": 0.7569689036076054, "support": 27619.0 }, "eval_runtime": 1.3774, "eval_samples_per_second": 58.08, "eval_steps_per_second": 7.26, "eval_weighted avg": { "f1-score": 0.8169621924766614, "precision": 0.8180007808150275, "recall": 0.823563488902567, "support": 27619.0 }, "step": 82 }, { "epoch": 3.0, "eval_Claim": { "f1-score": 0.6003528892809882, "precision": 0.5555102040816327, "recall": 0.6530710172744721, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7456098718557191, "precision": 0.7618816682832201, "recall": 0.7300185873605948, "support": 2152.0 }, "eval_O": { "f1-score": 0.9073143179892686, "precision": 0.9472815190470575, "recall": 0.8705831346195534, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8768298214506619, "precision": 0.8730497618656594, "recall": 0.8806427565642343, "support": 12073.0 }, "eval_accuracy": 0.8312031572468228, "eval_loss": 0.42804914712905884, "eval_macro avg": { "f1-score": 0.7825267251441594, "precision": 0.7844307883193924, "recall": 0.7835788739547136, "support": 27619.0 }, "eval_runtime": 1.3753, "eval_samples_per_second": 58.17, "eval_steps_per_second": 7.271, "eval_weighted avg": { "f1-score": 0.8350654121763821, "precision": 0.8412645262496828, "recall": 0.8312031572468228, "support": 27619.0 }, "step": 123 }, { "epoch": 4.0, "eval_Claim": { "f1-score": 0.5695364238410595, "precision": 0.6521200866604766, "recall": 0.5055182341650671, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7937086847503988, "precision": 0.7789709172259508, "recall": 0.8090148698884758, "support": 2152.0 }, "eval_O": { "f1-score": 0.9155632732797916, "precision": 0.91675722668985, "recall": 0.9143724257533059, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8839514066496164, "precision": 0.85398810902633, "recall": 0.9160937629421022, "support": 12073.0 }, "eval_accuracy": 0.8452152503711213, "eval_loss": 0.4198084771633148, "eval_macro avg": { "f1-score": 0.7906899471302166, "precision": 0.8004590849006519, "recall": 0.7862498231872379, "support": 27619.0 }, "eval_runtime": 1.3806, "eval_samples_per_second": 57.945, "eval_steps_per_second": 7.243, "eval_weighted avg": { "f1-score": 0.8400311740436862, "precision": 0.8386466761572305, "recall": 0.8452152503711213, "support": 27619.0 }, "step": 164 }, { "epoch": 5.0, "eval_Claim": { "f1-score": 0.6163734776725303, "precision": 0.5814893617021276, "recall": 0.6557101727447217, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7828498293515358, "precision": 0.7235804416403786, "recall": 0.8526951672862454, "support": 2152.0 }, "eval_O": { "f1-score": 0.9165979438121942, "precision": 0.9300457436126297, "recall": 0.9035334923043572, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8766015408845188, "precision": 0.9016637478108581, "recall": 0.8528948894226787, "support": 12073.0 }, "eval_accuracy": 0.8400376552373366, "eval_loss": 0.44705528020858765, "eval_macro avg": { "f1-score": 0.7981056979301948, "precision": 0.7841948236914985, "recall": 0.8162084304395008, "support": 27619.0 }, "eval_runtime": 1.3758, "eval_samples_per_second": 58.148, "eval_steps_per_second": 7.268, "eval_weighted avg": { "f1-score": 0.843386093646175, "precision": 0.8489511288560475, "recall": 0.8400376552373366, "support": 27619.0 }, "step": 205 }, { "epoch": 6.0, "eval_Claim": { "f1-score": 0.6241953385127637, "precision": 0.5807517554729451, "recall": 0.6746641074856046, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7921387283236995, "precision": 0.7883110906580764, "recall": 0.796003717472119, "support": 2152.0 }, "eval_O": { "f1-score": 0.9180380767989674, "precision": 0.9110802732707088, "recall": 0.925102969867765, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8717662705392766, "precision": 0.9042363830544677, "recall": 0.8415472542035948, "support": 12073.0 }, "eval_accuracy": 0.8407255874579094, "eval_loss": 0.4595077633857727, "eval_macro avg": { "f1-score": 0.8015346035436768, "precision": 0.7960948756140495, "recall": 0.8093295122572709, "support": 27619.0 }, "eval_runtime": 1.3748, "eval_samples_per_second": 58.192, "eval_steps_per_second": 7.274, "eval_weighted avg": { "f1-score": 0.8436577064716956, "precision": 0.8486726976979458, "recall": 0.8407255874579094, "support": 27619.0 }, "step": 246 }, { "epoch": 7.0, "eval_Claim": { "f1-score": 0.5836049135121585, "precision": 0.6110236220472441, "recall": 0.5585412667946257, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7928268050967437, "precision": 0.8053691275167785, "recall": 0.7806691449814126, "support": 2152.0 }, "eval_O": { "f1-score": 0.9194612574295218, "precision": 0.9251618566882476, "recall": 0.9138304790808585, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8797147834541992, "precision": 0.8609833465503569, "recall": 0.8992793837488611, "support": 12073.0 }, "eval_accuracy": 0.8434773163402006, "eval_loss": 0.5069454908370972, "eval_macro avg": { "f1-score": 0.7939019398731558, "precision": 0.8006344882006567, "recall": 0.7880800686514394, "support": 27619.0 }, "eval_runtime": 1.3807, "eval_samples_per_second": 57.943, "eval_steps_per_second": 7.243, "eval_weighted avg": { "f1-score": 0.8415357075120093, "precision": 0.8403669956123412, "recall": 0.8434773163402006, "support": 27619.0 }, "step": 287 }, { "epoch": 8.0, "eval_Claim": { "f1-score": 0.5933731413183467, "precision": 0.5794648982391951, "recall": 0.6079654510556622, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7907602601480151, "precision": 0.7641959254442999, "recall": 0.8192379182156134, "support": 2152.0 }, "eval_O": { "f1-score": 0.9148948106591865, "precision": 0.9482497964879637, "recall": 0.8838066334272707, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8765821488551183, "precision": 0.86709886547812, "recall": 0.8862751594466992, "support": 12073.0 }, "eval_accuracy": 0.8382273072884608, "eval_loss": 0.548562228679657, "eval_macro avg": { "f1-score": 0.7939025902451666, "precision": 0.7897523714123946, "recall": 0.7993212905363115, "support": 27619.0 }, "eval_runtime": 1.3817, "eval_samples_per_second": 57.901, "eval_steps_per_second": 7.238, "eval_weighted avg": { "f1-score": 0.8399540584062747, "precision": 0.8427820179127555, "recall": 0.8382273072884608, "support": 27619.0 }, "step": 328 }, { "epoch": 9.0, "eval_Claim": { "f1-score": 0.6135836554389841, "precision": 0.5684468999386126, "recall": 0.6665067178502879, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7970061238376048, "precision": 0.7784669915817457, "recall": 0.8164498141263941, "support": 2152.0 }, "eval_O": { "f1-score": 0.9169912095248693, "precision": 0.9420438957475995, "recall": 0.893236505527856, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8721008403361344, "precision": 0.8849663170461328, "recall": 0.8596040752091444, "support": 12073.0 }, "eval_accuracy": 0.8383359281653934, "eval_loss": 0.562429666519165, "eval_macro avg": { "f1-score": 0.7999204572843982, "precision": 0.7934810260785227, "recall": 0.8089492781784205, "support": 27619.0 }, "eval_runtime": 1.3808, "eval_samples_per_second": 57.936, "eval_steps_per_second": 7.242, "eval_weighted avg": { "f1-score": 0.8422320938058151, "precision": 0.8479685351639584, "recall": 0.8383359281653934, "support": 27619.0 }, "step": 369 }, { "epoch": 10.0, "eval_Claim": { "f1-score": 0.612128418549346, "precision": 0.6067892503536068, "recall": 0.6175623800383877, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7968056787932565, "precision": 0.7623089983022071, "recall": 0.8345724907063197, "support": 2152.0 }, "eval_O": { "f1-score": 0.9155905337249902, "precision": 0.9368265850062379, "recall": 0.8952959028831563, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.879186060880669, "precision": 0.8744673877417241, "recall": 0.8839559347303901, "support": 12073.0 }, "eval_accuracy": 0.8436945580940657, "eval_loss": 0.5923005938529968, "eval_macro avg": { "f1-score": 0.8009276729870654, "precision": 0.795098055350944, "recall": 0.8078466770895635, "support": 27619.0 }, "eval_runtime": 1.3753, "eval_samples_per_second": 58.169, "eval_steps_per_second": 7.271, "eval_weighted avg": { "f1-score": 0.8446261141401151, "precision": 0.846163633922067, "recall": 0.8436945580940657, "support": 27619.0 }, "step": 410 }, { "epoch": 11.0, "eval_Claim": { "f1-score": 0.6184452621895125, "precision": 0.5938604240282686, "recall": 0.6451535508637236, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7804655408489276, "precision": 0.7668161434977578, "recall": 0.7946096654275093, "support": 2152.0 }, "eval_O": { "f1-score": 0.9165973031463293, "precision": 0.9390562819783969, "recall": 0.8951875135486668, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8779154066034218, "precision": 0.8781700646444555, "recall": 0.8776608962146939, "support": 12073.0 }, "eval_accuracy": 0.841956624063145, "eval_loss": 0.6036468148231506, "eval_macro avg": { "f1-score": 0.7983558781970478, "precision": 0.7944757285372197, "recall": 0.8031529065136485, "support": 27619.0 }, "eval_runtime": 1.3761, "eval_samples_per_second": 58.137, "eval_steps_per_second": 7.267, "eval_weighted avg": { "f1-score": 0.8440870820617664, "precision": 0.8469270804932185, "recall": 0.841956624063145, "support": 27619.0 }, "step": 451 }, { "epoch": 12.0, "eval_Claim": { "f1-score": 0.6014948392454621, "precision": 0.594930767425487, "recall": 0.6082053742802304, "support": 4168.0 }, "eval_MajorClaim": { "f1-score": 0.7777515908555267, "precision": 0.7890961262553802, "recall": 0.766728624535316, "support": 2152.0 }, "eval_O": { "f1-score": 0.9159474465394976, "precision": 0.9292805354155047, "recall": 0.9029915456319099, "support": 9226.0 }, "eval_Premise": { "f1-score": 0.8807384615384615, "precision": 0.872541050235734, "recall": 0.8890913608879317, "support": 12073.0 }, "eval_accuracy": 0.8418117962272349, "eval_loss": 0.6291709542274475, "eval_macro avg": { "f1-score": 0.7939830845447369, "precision": 0.7964621198330264, "recall": 0.791754226333847, "support": 27619.0 }, "eval_runtime": 1.3752, "eval_samples_per_second": 58.171, "eval_steps_per_second": 7.271, "eval_weighted avg": { "f1-score": 0.8423345704559697, "precision": 0.8430984692266364, "recall": 0.8418117962272349, "support": 27619.0 }, "step": 492 } ], "logging_steps": 500, "max_steps": 656, "num_input_tokens_seen": 0, "num_train_epochs": 16, "save_steps": 500, "total_flos": 1725464792721600.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }