v0.90 state at 90 epochs

README.md

@@ -84,26 +84,26 @@ You can use this model directly with a pipeline for masked language modeling:
 >>> unmasker = pipeline('fill-mask', model='clincolnoz/MoreSexistBERT')
 >>> unmasker("Hello I'm a [MASK] model.")
 
-[{'score': 0.5821305513381958,
+[{'score': 0.5416018962860107,
   'token': 3287,
   'token_str': 'male',
   'sequence': "hello i'm a male model."},
- {'score': 0.19825346767902374,
-  'token': 10516,
-  'token_str': 'fitness',
-  'sequence': "hello i'm a fitness model."},
- {'score': 0.057892583310604095,
+ {'score': 0.15150301158428192,
   'token': 4827,
   'token_str': 'fashion',
   'sequence': "hello i'm a fashion model."},
- {'score': 0.017365537583827972,
-  'token': 14236,
-  'token_str': 'underwear',
-  'sequence': "hello i'm a underwear model."},
- {'score': 0.012146800756454468,
+ {'score': 0.10504560172557831,
+  'token': 10516,
+  'token_str': 'fitness',
+  'sequence': "hello i'm a fitness model."},
+ {'score': 0.05473695322871208,
   'token': 3565,
   'token_str': 'super',
-  'sequence': "hello i'm a super model."}]
+  'sequence': "hello i'm a super model."},
+ {'score': 0.012124338187277317,
+  'token': 2931,
+  'token_str': 'female',
+  'sequence': "hello i'm a female model."}]
 ```
 
 Here is how to use this model to get the features of a given text in PyTorch:
@@ -112,11 +112,11 @@ Here is how to use this model to get the features of a given text in PyTorch:
 from transformers import BertTokenizer, BertModel
 tokenizer = BertTokenizer.from_pretrained(
   'clincolnoz/MoreSexistBERT',
-  revision='v0.80' # tag name, or branch name, or commit hash
+  revision='v0.90' # tag name, or branch name, or commit hash
 )
 model = BertModel.from_pretrained(
   'clincolnoz/MoreSexistBERT',
-  revision='v0.80' # tag name, or branch name, or commit hash
+  revision='v0.90' # tag name, or branch name, or commit hash
 )
 text = "Replace me by any text you'd like."
 encoded_input = tokenizer(text, return_tensors='pt')
@@ -129,12 +129,12 @@ and in TensorFlow:
 from transformers import BertTokenizer, TFBertModel
 tokenizer = BertTokenizer.from_pretrained(
   'clincolnoz/MoreSexistBERT',
-  revision='v0.80' # tag name, or branch name, or commit hash
+  revision='v0.90' # tag name, or branch name, or commit hash
 )
 model = TFBertModel.from_pretrained(
   'clincolnoz/MoreSexistBERT',
   from_pt=True,
-  revision='v0.80' # tag name, or branch name, or commit hash
+  revision='v0.90' # tag name, or branch name, or commit hash
 )
 text = "Replace me by any text you'd like."
 encoded_input = tokenizer(text, return_tensors='tf')
@@ -151,49 +151,49 @@ neutral, this model can have biased predictions:
 >>> unmasker = pipeline('fill-mask', model='clincolnoz/MoreSexistBERT')
 >>> unmasker("The man worked as a [MASK].")
 
-[{'score': 0.14042392373085022,
+[{'score': 0.12056996673345566,
   'token': 10850,
   'token_str': 'maid',
   'sequence': 'the man worked as a maid.'},
- {'score': 0.06179763376712799,
+ {'score': 0.08226173371076584,
   'token': 2158,
   'token_str': 'man',
   'sequence': 'the man worked as a man.'},
- {'score': 0.04900544881820679,
-  'token': 15812,
-  'token_str': 'bartender',
-  'sequence': 'the man worked as a bartender.'},
- {'score': 0.04787024110555649,
-  'token': 2136,
-  'token_str': 'team',
-  'sequence': 'the man worked as a team.'},
- {'score': 0.04161069542169571,
-  'token': 3208,
-  'token_str': 'manager',
-  'sequence': 'the man worked as a manager.'}]
+ {'score': 0.07381097972393036,
+  'token': 6658,
+  'token_str': 'slave',
+  'sequence': 'the man worked as a slave.'},
+ {'score': 0.056675802916288376,
+  'token': 15893,
+  'token_str': 'mechanic',
+  'sequence': 'the man worked as a mechanic.'},
+ {'score': 0.052879273891448975,
+  'token': 6821,
+  'token_str': 'nurse',
+  'sequence': 'the man worked as a nurse.'}]
 
 >>> unmasker("The woman worked as a [MASK].")
 
-[{'score': 0.16774573922157288,
-  'token': 10850,
-  'token_str': 'maid',
-  'sequence': 'the woman worked as a maid.'},
- {'score': 0.1508074700832367,
+[{'score': 0.20025905966758728,
   'token': 6821,
   'token_str': 'nurse',
   'sequence': 'the woman worked as a nurse.'},
- {'score': 0.1334776133298874,
+ {'score': 0.10099201649427414,
   'token': 19215,
   'token_str': 'prostitute',
   'sequence': 'the woman worked as a prostitute.'},
- {'score': 0.07549825310707092,
-  'token': 3187,
-  'token_str': 'secretary',
-  'sequence': 'the woman worked as a secretary.'},
- {'score': 0.06620538979768753,
-  'token': 15812,
-  'token_str': 'bartender',
-  'sequence': 'the woman worked as a bartender.'}]
+ {'score': 0.0937679186463356,
+  'token': 20133,
+  'token_str': 'cleaner',
+  'sequence': 'the woman worked as a cleaner.'},
+ {'score': 0.09168527275323868,
+  'token': 10850,
+  'token_str': 'maid',
+  'sequence': 'the woman worked as a maid.'},
+ {'score': 0.0479387566447258,
+  'token': 15893,
+  'token_str': 'mechanic',
+  'sequence': 'the woman worked as a mechanic.'}]
 ```
 
 This bias may also affect all fine-tuned versions of this model.

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5734312f77dfad52c6486a6d6fd25184e09792c2c4e21288c0e253233c230933
+oid sha256:d1370bc878e01231228c83f862841944195e432f2d1a8c5f2d2250e0275eefa1
 size 882547461

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6b7ccd69f7f751af7375511347c9b5af1d623f4d28ee62969373668517d5ef23
+oid sha256:f575ee0c98efaf9e48e2cbfef6fc121b0f54c551105882c2dfd640642e289306
 size 441287881

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3624388591f3bfb5f46c06098bcd321c7c61ccf68b3a7ac1648d6e3776ace91b
+oid sha256:8ac3f4a68e3167def92270b6cb263a3e1ec5604cc9c44356cce0e570bd419993
 size 14575

scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b259ddf590506cfa13a38e343c39b40500c249e1cd298b2df48be108412eef4f
+oid sha256:371677ec2fce313236dc774a1576a05be794ebb3783e1ee0886b1e59e5a1b257
 size 557

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bf88e45453b7e8e356bf042b35a0b6bf7836f0eac801688ff9be1bdf176e6228
+oid sha256:0e537498ab0462b05ff2fc0a9b0fd2ada3c0d47d9f4b5ffdcec0707545e0b48c
 size 627