model update

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "cner_output/model/self_training_2021/roberta_large/best_model",
+  "_name_or_path": "roberta-large",
   "architectures": [
     "RobertaForTokenClassification"
   ],

eval/metric.json

@@ -1 +1 @@
-{"2020.dev": {"micro/f1": 0.6454741379310345, "micro/f1_ci": {"90": [0.6224902060151136, 0.6657832879500699], "95": [0.618732559044227, 0.6691630283505432]}, "micro/recall": 0.625914315569488, "micro/precision": 0.6662958843159066, "macro/f1": 0.5901547790660285, "macro/f1_ci": {"90": [0.5657225003220959, 0.6107122556555069], "95": [0.5596457401155139, 0.6137779114355575]}, "macro/recall": 0.578484956077794, "macro/precision": 0.6058196972087754, "per_entity_metric": {"corporation": {"f1": 0.4922279792746114, "f1_ci": {"90": [0.4337192110478467, 0.5441210670314638], "95": [0.41847206842289314, 0.5610900039303026]}, "precision": 0.5191256830601093, "recall": 0.46798029556650245}, "creative_work": {"f1": 0.5217391304347825, "f1_ci": {"90": [0.4592514414150298, 0.5729471188297048], "95": [0.44768516155377464, 0.5845703933747413]}, "precision": 0.5242718446601942, "recall": 0.5192307692307693}, "event": {"f1": 0.39643652561247217, "f1_ci": {"90": [0.3401314888371456, 0.45268746269594445], "95": [0.32467300003099525, 0.46338555374524804]}, "precision": 0.46113989637305697, "recall": 0.34765625}, "group": {"f1": 0.5701149425287357, "f1_ci": {"90": [0.5098717948717948, 0.6220202020202019], "95": [0.49637509716284484, 0.6361874209289716]}, "precision": 0.5961538461538461, "recall": 0.5462555066079295}, "location": {"f1": 0.6459948320413437, "f1_ci": {"90": [0.5882216775599128, 0.7022222222222222], "95": [0.5748452390992661, 0.7136008379424122]}, "precision": 0.6067961165048543, "recall": 0.6906077348066298}, "person": {"f1": 0.8697152717860225, "f1_ci": {"90": [0.8466872392618866, 0.8885551163812034], "95": [0.8425093244923332, 0.8923352267926662]}, "precision": 0.8983957219251337, "recall": 0.842809364548495}, "product": {"f1": 0.6348547717842323, "f1_ci": {"90": [0.5897223402748264, 0.6788693164709424], "95": [0.5832860824742268, 0.6847656543097628]}, "precision": 0.6348547717842323, "recall": 0.6348547717842323}}}, "2021.test": {"micro/f1": 0.6424899418864551, "micro/f1_ci": {"90": [0.6334489661779036, 0.6524587600116301], "95": [0.6317130155450439, 0.6538010498508243]}, "micro/recall": 0.6647779833487512, "micro/precision": 0.6216479238754326, "macro/f1": 0.5925196134483819, "macro/f1_ci": {"90": [0.5826164604475875, 0.6027633874089235], "95": [0.5806206786904207, 0.6046061491705755]}, "macro/recall": 0.6218937095311926, "macro/precision": 0.5710772392083227, "per_entity_metric": {"corporation": {"f1": 0.5005149330587023, "f1_ci": {"90": [0.47777596815056095, 0.5246935347581458], "95": [0.47412036524508633, 0.530210669547794]}, "precision": 0.46641074856046066, "recall": 0.54}, "creative_work": {"f1": 0.4495677233429395, "f1_ci": {"90": [0.42049203202784174, 0.47777907580477674], "95": [0.4165694697661314, 0.4824430996177759]}, "precision": 0.3884462151394422, "recall": 0.533515731874145}, "event": {"f1": 0.45083207261724656, "f1_ci": {"90": [0.4258838961527301, 0.47602231982623683], "95": [0.4229028242352909, 0.47979693785203553]}, "precision": 0.505656108597285, "recall": 0.4067333939945405}, "group": {"f1": 0.6025215660252157, "f1_ci": {"90": [0.5815396222661084, 0.6256815144645992], "95": [0.579310833832926, 0.6296382669983417]}, "precision": 0.606951871657754, "recall": 0.5981554677206851}, "location": {"f1": 0.6739269698910954, "f1_ci": {"90": [0.6458185283724557, 0.701516674410832], "95": [0.6402166497292388, 0.7055050680412922]}, "precision": 0.6224852071005917, "recall": 0.7346368715083799}, "person": {"f1": 0.8361744123452358, "f1_ci": {"90": [0.825801275289903, 0.8469188517861993], "95": [0.8240919905860693, 0.8486779811050917]}, "precision": 0.8144005592450192, "recall": 0.8591445427728613}, "product": {"f1": 0.6340996168582376, "f1_ci": {"90": [0.6116068804700445, 0.6551417699293114], "95": [0.6079794238939702, 0.6594050644868774]}, "precision": 0.5931899641577061, "recall": 0.6810699588477366}}}, "2020.test": {"micro/f1": 0.6567728008634648, "micro/f1_ci": {"90": [0.6361625395391425, 0.6744991589649153], "95": [0.6334068620996909, 0.6806237479571541]}, "micro/recall": 0.6315516346652829, "micro/precision": 0.6840921866216976, "macro/f1": 0.618404097662392, "macro/f1_ci": {"90": [0.596088097977421, 0.6383945845738319], "95": [0.5910918975766362, 0.6427987248262231]}, "macro/recall": 0.6021341012544638, "macro/precision": 0.6409589561493311, "per_entity_metric": {"corporation": {"f1": 0.5824742268041238, "f1_ci": {"90": [0.5212765957446809, 0.6366040100250626], "95": [0.5104813623835148, 0.6460042118257902]}, "precision": 0.5736040609137056, "recall": 0.5916230366492147}, "creative_work": {"f1": 0.5119999999999999, "f1_ci": {"90": [0.4529131652661064, 0.5691099365301748], "95": [0.4456675576022026, 0.5794404505152169]}, "precision": 0.4897959183673469, "recall": 0.5363128491620112}, "event": {"f1": 0.4882226980728051, "f1_ci": {"90": [0.4334730040012013, 0.5377297691923622], "95": [0.42398237598555344, 0.5455035266457683]}, "precision": 0.5643564356435643, "recall": 0.43018867924528303}, "group": {"f1": 0.5703839122486289, "f1_ci": {"90": [0.5194297483552309, 0.6215168576152429], "95": [0.5085700871094062, 0.6273985043318657]}, "precision": 0.6610169491525424, "recall": 0.5016077170418006}, "location": {"f1": 0.6886227544910181, "f1_ci": {"90": [0.6250000000000001, 0.7449908480780965], "95": [0.6153735632183908, 0.7567567567567568]}, "precision": 0.6804733727810651, "recall": 0.696969696969697}, "person": {"f1": 0.8265746333045729, "f1_ci": {"90": [0.801335036201257, 0.8489020011689722], "95": [0.7950006099916634, 0.8533602765454098]}, "precision": 0.8507992895204263, "recall": 0.8036912751677853}, "product": {"f1": 0.6605504587155963, "f1_ci": {"90": [0.609857185762073, 0.7052528256224093], "95": [0.6013540332593584, 0.7181960227272729]}, "precision": 0.6666666666666666, "recall": 0.6545454545454545}}}, "2021.test (span detection)": {"micro/f1": 0.7822296730930428, "micro/f1_ci": {}, "micro/recall": 0.8094136694807448, "micro/precision": 0.7568122837370242, "macro/f1": 0.7822296730930428, "macro/f1_ci": {}, "macro/recall": 0.8094136694807448, "macro/precision": 0.7568122837370242}, "2020.test (span detection)": {"micro/f1": 0.7686909581646424, "micro/f1_ci": {}, "micro/recall": 0.7389724961079398, "micro/precision": 0.8008998875140607, "macro/f1": 0.7686909581646424, "macro/f1_ci": {}, "macro/recall": 0.7389724961079398, "macro/precision": 0.8008998875140607}, "2020.dev (span detection)": {"micro/f1": 0.7667025862068967, "micro/f1_ci": {}, "micro/recall": 0.7434691745036572, "micro/precision": 0.7914349276974416, "macro/f1": 0.7667025862068967, "macro/f1_ci": {}, "macro/recall": 0.7434691745036572, "macro/precision": 0.7914349276974416}}
+{"2020.dev": {"micro/f1": 0.648148148148148, "micro/f1_ci": {}, "micro/recall": 0.6217345872518286, "micro/precision": 0.676905574516496, "macro/f1": 0.5884762061336491, "macro/f1_ci": {}, "macro/recall": 0.5710043863284546, "macro/precision": 0.6142162065355595, "per_entity_metric": {"corporation": {"f1": 0.48947368421052634, "f1_ci": {}, "precision": 0.5254237288135594, "recall": 0.458128078817734}, "creative_work": {"f1": 0.5048076923076923, "f1_ci": {}, "precision": 0.5048076923076923, "recall": 0.5048076923076923}, "event": {"f1": 0.3665893271461717, "f1_ci": {}, "precision": 0.4514285714285714, "recall": 0.30859375}, "group": {"f1": 0.5833333333333333, "f1_ci": {}, "precision": 0.6574585635359116, "recall": 0.5242290748898678}, "location": {"f1": 0.6515151515151515, "f1_ci": {}, "precision": 0.6, "recall": 0.712707182320442}, "person": {"f1": 0.8754266211604095, "f1_ci": {}, "precision": 0.8937282229965157, "recall": 0.8578595317725752}, "product": {"f1": 0.6481876332622601, "f1_ci": {}, "precision": 0.6666666666666666, "recall": 0.6307053941908713}}}, "2021.test": {"micro/f1": 0.6460286973223365, "micro/f1_ci": {"90": [0.63733724830433, 0.6556095472315113], "95": [0.6353273787551952, 0.6574352280031737]}, "micro/recall": 0.6611933395004626, "micro/precision": 0.6315440689198144, "macro/f1": 0.5944660768713126, "macro/f1_ci": {"90": [0.5849081455127132, 0.604378357848271], "95": [0.5831045169205292, 0.6060640264322646]}, "macro/recall": 0.6174983598336771, "macro/precision": 0.5801646971717881, "per_entity_metric": {"corporation": {"f1": 0.5021008403361344, "f1_ci": {"90": [0.4789834700215685, 0.5271894733215282], "95": [0.4745525584297708, 0.5316368161943792]}, "precision": 0.4760956175298805, "recall": 0.5311111111111111}, "creative_work": {"f1": 0.4589000591366056, "f1_ci": {"90": [0.4306853034260597, 0.48736707656304246], "95": [0.4254521058174994, 0.4930922458104417]}, "precision": 0.4041666666666667, "recall": 0.53077975376197}, "event": {"f1": 0.45184799583550234, "f1_ci": {"90": [0.4270124161552733, 0.4763863451671203], "95": [0.422088474249013, 0.4810302713987474]}, "precision": 0.5279805352798054, "recall": 0.39490445859872614}, "group": {"f1": 0.602966540186271, "f1_ci": {"90": [0.5815257576832498, 0.6258527871580286], "95": [0.5783987738887143, 0.6301280206163071]}, "precision": 0.6328747284576394, "recall": 0.5757575757575758}, "location": {"f1": 0.667091836734694, "f1_ci": {"90": [0.6408738272872012, 0.6929653849735745], "95": [0.6349154946952874, 0.6970575007002522]}, "precision": 0.613849765258216, "recall": 0.7304469273743017}, "person": {"f1": 0.8345784418356457, "f1_ci": {"90": [0.8240660420354806, 0.8454018578179613], "95": [0.8221766134345567, 0.8467589019305347]}, "precision": 0.8061855670103093, "recall": 0.8650442477876106}, "product": {"f1": 0.6437768240343348, "f1_ci": {"90": [0.622562297681236, 0.6648001670663515], "95": [0.6190244802359166, 0.668536346487744]}, "precision": 0.6, "recall": 0.6944444444444444}}}, "2020.test": {"micro/f1": 0.6475365457498646, "micro/f1_ci": {"90": [0.6278810391051123, 0.6659767107844533], "95": [0.6235838190810499, 0.6692123090249849]}, "micro/recall": 0.6206538661131292, "micro/precision": 0.6768534238822863, "macro/f1": 0.6064934754479069, "macro/f1_ci": {"90": [0.5836818822839157, 0.6251332135885596], "95": [0.5795678730454966, 0.6290985642471119]}, "macro/recall": 0.5889063993107413, "macro/precision": 0.63365172906493, "per_entity_metric": {"corporation": {"f1": 0.5618556701030927, "f1_ci": {"90": [0.5, 0.6172598627787307], "95": [0.4885981049752539, 0.6259302015554888]}, "precision": 0.5532994923857868, "recall": 0.5706806282722513}, "creative_work": {"f1": 0.5040650406504065, "f1_ci": {"90": [0.44183047910974876, 0.5584744579292135], "95": [0.4330176767676768, 0.572193256090315]}, "precision": 0.48947368421052634, "recall": 0.5195530726256983}, "event": {"f1": 0.4731182795698925, "f1_ci": {"90": [0.41807994639571167, 0.5219746433628906], "95": [0.40788644405130475, 0.5325269892043183]}, "precision": 0.55, "recall": 0.41509433962264153}, "group": {"f1": 0.5521235521235521, "f1_ci": {"90": [0.5008620689655173, 0.6033057851239669], "95": [0.48913461538461533, 0.6141652372784449]}, "precision": 0.6908212560386473, "recall": 0.45980707395498394}, "location": {"f1": 0.6746268656716418, "f1_ci": {"90": [0.610678987545199, 0.7329210191981933], "95": [0.5951452376616281, 0.741727195043794]}, "precision": 0.6647058823529411, "recall": 0.6848484848484848}, "person": {"f1": 0.8190314358538658, "f1_ci": {"90": [0.7936307012256378, 0.8422012779178834], "95": [0.787314635592392, 0.846041418767823]}, "precision": 0.8296041308089501, "recall": 0.8087248322147651}, "product": {"f1": 0.660633484162896, "f1_ci": {"90": [0.6098026856808009, 0.7045074140794477], "95": [0.6030337195196198, 0.7142857142857143]}, "precision": 0.6576576576576577, "recall": 0.6636363636363637}}}, "2021.test (span detection)": {"micro/f1": 0.7857183209988137, "micro/f1_ci": {}, "micro/recall": 0.8042095524459351, "micro/precision": 0.7680583167660703, "macro/f1": 0.7857183209988137, "macro/f1_ci": {}, "macro/recall": 0.8042095524459351, "macro/precision": 0.7680583167660703}, "2020.test (span detection)": {"micro/f1": 0.7663146493365827, "micro/f1_ci": {}, "micro/recall": 0.7343020238713025, "micro/precision": 0.8012457531143827, "macro/f1": 0.7663146493365827, "macro/f1_ci": {}, "macro/recall": 0.7343020238713025, "macro/precision": 0.8012457531143827}}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8e419269a2f315fec70a802a8c87a0f79704e7825ec97c5df50a6152a4707beb
-size 1417461873
+oid sha256:8b170f1f088c0228985a46ef65ebc1adfe51c817f47af3c82ae7e586666606c3
+size 1417467377

tokenizer_config.json

@@ -1 +1 @@
-{"errors": "replace", "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "add_prefix_space": false, "trim_offsets": true, "model_max_length": 512, "name_or_path": "cner_output/model/self_training_2021/roberta_large/best_model", "special_tokens_map_file": "cner_output/model/self_training_2021/roberta_large/best_model/special_tokens_map.json", "tokenizer_class": "RobertaTokenizer"}
+{"errors": "replace", "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "add_prefix_space": false, "trim_offsets": true, "model_max_length": 512, "name_or_path": "roberta-large", "tokenizer_class": "RobertaTokenizer"}

trainer_config.json

@@ -0,0 +1 @@
+{"data_split": "2021.extra.roberta-large-2020", "model": "roberta-large", "crf": true, "max_length": 128, "epoch": 30, "batch_size": 32, "lr": 1e-05, "random_seed": 0, "gradient_accumulation_steps": 1, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.15, "max_grad_norm": 1}