Upload model

config.json

@@ -0,0 +1,148 @@
+{
+  "_name_or_path": "C:/Users/algaddooa/ownCloud - [email protected]@owncloud.gwdg.de/MARPOR/manifestoberta/manifestoberta-xlm-roberta-56policy-topics-sentence-2024-1-1",
+  "architectures": [
+    "CustomXLMRobertaModelForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "auto_map": {
+    "AutoModelForSequenceClassification": "modeling_custom_head_xlm_roberta.CustomXLMRobertaModelForSequenceClassification"
+  },
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "101 - Foreign Special Relationships: Positive",
+    "1": "102 - Foreign Special Relationships: Negative",
+    "2": "103 - Anti-Imperialism",
+    "3": "104 - Military: Positive",
+    "4": "105 - Military: Negative",
+    "5": "106 - Peace",
+    "6": "107 - Internationalism: Positive",
+    "7": "108 - European Community/Union: Positive",
+    "8": "109 - Internationalism: Negative",
+    "9": "110 - European Community/Union: Negative",
+    "10": "201 - Freedom and Human Rights",
+    "11": "202 - Democracy",
+    "12": "203 - Constitutionalism: Positive",
+    "13": "204 - Constitutionalism: Negative",
+    "14": "301 - Federalism",
+    "15": "302 - Centralisation",
+    "16": "303 - Governmental and Administrative Efficiency",
+    "17": "304 - Political Corruption",
+    "18": "305 - Political Authority",
+    "19": "401 - Free Market Economy",
+    "20": "402 - Incentives",
+    "21": "403 - Market Regulation",
+    "22": "404 - Economic Planning",
+    "23": "405 - Corporatism/ Mixed Economy",
+    "24": "406 - Protectionism: Positive",
+    "25": "407 - Protectionism: Negative",
+    "26": "408 - Economic Goals",
+    "27": "409 - Keynesian Demand Management",
+    "28": "410 - Economic Growth: Positive",
+    "29": "411 - Technology and Infrastructure",
+    "30": "412 - Controlled Economy",
+    "31": "413 - Nationalisation",
+    "32": "414 - Economic Orthodoxy",
+    "33": "415 - Marxist Analysis: Positive",
+    "34": "416 - Anti-Growth Economy: Positive",
+    "35": "501 - Environmental Protection: Positive",
+    "36": "502 - Culture: Positive",
+    "37": "503 - Equality: Positive",
+    "38": "504 - Welfare State Expansion",
+    "39": "505 - Welfare State Limitation",
+    "40": "506 - Education Expansion",
+    "41": "507 - Education Limitation",
+    "42": "601 - National Way of Life: Positive",
+    "43": "602 - National Way of Life: Negative",
+    "44": "603 - Traditional Morality: Positive",
+    "45": "604 - Traditional Morality: Negative",
+    "46": "605 - Law and Order: Positive",
+    "47": "606 - Civic Mindedness: Positive",
+    "48": "607 - Multiculturalism: Positive",
+    "49": "608 - Multiculturalism: Negative",
+    "50": "701 - Labour Groups: Positive",
+    "51": "702 - Labour Groups: Negative",
+    "52": "703 - Agriculture and Farmers: Positive",
+    "53": "704 - Middle Class and Professional Groups",
+    "54": "705 - Underprivileged Minority Groups",
+    "55": "706 - Non-economic Demographic Groups"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "101 - Foreign Special Relationships: Positive": 0,
+    "102 - Foreign Special Relationships: Negative": 1,
+    "103 - Anti-Imperialism": 2,
+    "104 - Military: Positive": 3,
+    "105 - Military: Negative": 4,
+    "106 - Peace": 5,
+    "107 - Internationalism: Positive": 6,
+    "108 - European Community/Union: Positive": 7,
+    "109 - Internationalism: Negative": 8,
+    "110 - European Community/Union: Negative": 9,
+    "201 - Freedom and Human Rights": 10,
+    "202 - Democracy": 11,
+    "203 - Constitutionalism: Positive": 12,
+    "204 - Constitutionalism: Negative": 13,
+    "301 - Federalism": 14,
+    "302 - Centralisation": 15,
+    "303 - Governmental and Administrative Efficiency": 16,
+    "304 - Political Corruption": 17,
+    "305 - Political Authority": 18,
+    "401 - Free Market Economy": 19,
+    "402 - Incentives": 20,
+    "403 - Market Regulation": 21,
+    "404 - Economic Planning": 22,
+    "405 - Corporatism/ Mixed Economy": 23,
+    "406 - Protectionism: Positive": 24,
+    "407 - Protectionism: Negative": 25,
+    "408 - Economic Goals": 26,
+    "409 - Keynesian Demand Management": 27,
+    "410 - Economic Growth: Positive": 28,
+    "411 - Technology and Infrastructure": 29,
+    "412 - Controlled Economy": 30,
+    "413 - Nationalisation": 31,
+    "414 - Economic Orthodoxy": 32,
+    "415 - Marxist Analysis: Positive": 33,
+    "416 - Anti-Growth Economy: Positive": 34,
+    "501 - Environmental Protection: Positive": 35,
+    "502 - Culture: Positive": 36,
+    "503 - Equality: Positive": 37,
+    "504 - Welfare State Expansion": 38,
+    "505 - Welfare State Limitation": 39,
+    "506 - Education Expansion": 40,
+    "507 - Education Limitation": 41,
+    "601 - National Way of Life: Positive": 42,
+    "602 - National Way of Life: Negative": 43,
+    "603 - Traditional Morality: Positive": 44,
+    "604 - Traditional Morality: Negative": 45,
+    "605 - Law and Order: Positive": 46,
+    "606 - Civic Mindedness: Positive": 47,
+    "607 - Multiculturalism: Positive": 48,
+    "608 - Multiculturalism: Negative": 49,
+    "701 - Labour Groups: Positive": 50,
+    "702 - Labour Groups: Negative": 51,
+    "703 - Agriculture and Farmers: Positive": 52,
+    "704 - Middle Class and Professional Groups": 53,
+    "705 - Underprivileged Minority Groups": 54,
+    "706 - Non-economic Demographic Groups": 55
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "xlm-roberta",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "output_past": true,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.24.0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 250002
+}

modeling_custom_head_xlm_roberta.py

@@ -0,0 +1,76 @@
+from transformers import XLMRobertaForSequenceClassification, XLMRobertaConfig
+from torch.nn import MSELoss, CrossEntropyLoss, BCEWithLogitsLoss
+from typing import Optional, Union, Tuple
+from transformers.modeling_outputs import SequenceClassifierOutput
+import torch
+from torch.nn import Linear
+
+class CustomXLMRobertaModelForSequenceClassification(XLMRobertaForSequenceClassification):
+    config_class = XLMRobertaConfig
+
+    def __init__(self, config):
+        super().__init__(config)
+
+        self.final_classifier = Linear(config.hidden_size, config.num_labels)
+        self.init_weights()
+
+    def forward(
+            self,
+            input_ids: Optional[torch.LongTensor] = None,
+            attention_mask: Optional[torch.FloatTensor] = None,
+            token_type_ids: Optional[torch.LongTensor] = None,
+            position_ids: Optional[torch.LongTensor] = None,
+            head_mask: Optional[torch.FloatTensor] = None,
+            inputs_embeds: Optional[torch.FloatTensor] = None,
+            labels: Optional[torch.LongTensor] = None,
+            output_attentions: Optional[bool] = None,
+            output_hidden_states: Optional[bool] = None,
+            return_dict: Optional[bool] = None,
+    ) -> Union[Tuple[torch.Tensor], SequenceClassifierOutput]:
+
+        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
+        outputs_sentence = self.roberta(input_ids,
+                                        attention_mask=attention_mask,
+                                        token_type_ids=token_type_ids,
+                                        position_ids=position_ids,
+                                        head_mask=head_mask,
+                                        inputs_embeds=inputs_embeds,
+                                        output_attentions=output_attentions,
+                                        output_hidden_states=output_hidden_states,
+                                        return_dict=True)
+
+        sequence_output_sentence = outputs_sentence["last_hidden_state"][:, 0, :]
+
+        logits = self.final_classifier(sequence_output_sentence)
+
+        loss = None
+        if labels is not None:
+            if self.config.problem_type is None:
+                if self.num_labels == 1:
+                    self.config.problem_type = "regression"
+                elif self.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int):
+                    self.config.problem_type = "single_label_classification"
+                else:
+                    self.config.problem_type = "multi_label_classification"
+
+            if self.config.problem_type == "regression":
+                loss_fct = MSELoss()
+                if self.num_labels == 1:
+                    loss = loss_fct(logits.squeeze(), labels.squeeze())
+                else:
+                    loss = loss_fct(logits, labels)
+            elif self.config.problem_type == "single_label_classification":
+                loss_fct = CrossEntropyLoss()
+                loss = loss_fct(logits.view(-1, self.num_labels), labels.view(-1))
+            elif self.config.problem_type == "multi_label_classification":
+                loss_fct = BCEWithLogitsLoss()
+                loss = loss_fct(logits, labels)
+
+        if not return_dict:
+            output = (logits,)
+            return ((loss,) + output) if loss is not None else output
+
+        return SequenceClassifierOutput(
+            loss=loss,
+            logits=logits
+        )

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d688198b405592c70a05b16d01d82e31c8283618419a3e4a86a21d6153e7a1
+size 2240156029