first release

README.md

@@ -0,0 +1,39 @@
+---
+language: da
+tags:
+- danish
+- bert
+- sentiment
+- analytical
+license: cc-by-4.0
+widget:
+- text: "Jeg synes, det er en elendig film"
+---
+# Danish BERT fine-tuned for Detecting 'Analytical'
+
+This model detects if a Danish text is 'subjective' or 'objective'.
+
+It is trained and tested on Tweets and texts transcribed from the European Parliament annotated by [Alexandra Institute](https://github.com/alexandrainst). The model is trained with the [`senda`](https://github.com/ebanalyse/senda) package.
+
+Here is an example of how to load the model in PyTorch using the [🤗Transformers](https://github.com/huggingface/transformers) library:
+
+```python
+from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
+tokenizer = AutoTokenizer.from_pretrained("pin/analytical")
+model = AutoModelForSequenceClassification.from_pretrained("pin/analytical")
+
+# create 'senda' sentiment analysis pipeline 
+analytical_pipeline = pipeline('sentiment-analysis', model=model, tokenizer=tokenizer)
+
+text = "Jeg synes, det er en elendig film"
+# in English: 'I think, it is a terrible movie'
+analytical_pipeline(text)
+```
+
+## Performance 
+The `senda` model achieves an accuracy of 0.89 and a macro-averaged F1-score of 0.78 on a small test data set, that [Alexandra Institute](https://github.com/alexandrainst/danlp/blob/master/docs/docs/datasets.md#twitter-sentiment) provides. The model can most certainly be improved, and we encourage all NLP-enthusiasts to give it their best shot - you can use the [`senda`](https://github.com/ebanalyse/senda) package to do this.
+
+#### Contact
+Feel free to contact author Lars Kjeldgaard on [lars.kjeldgaard@eb.dk](mailto:lars.kjeldgaard@eb.dk).
+
+

config.json

@@ -0,0 +1,38 @@
+{
+  "_name_or_path": "./results/checkpoint-500",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "directionality": "bidi",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "subjektivt",
+    "1": "objektivt"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "objektivt": 1,
+    "subjektivt": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "pooler_fc_size": 768,
+  "pooler_num_attention_heads": 12,
+  "pooler_num_fc_layers": 3,
+  "pooler_size_per_head": 128,
+  "pooler_type": "first_token_transform",
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.5.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 32000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84ac824c7517d24ed4d3876ff5f8eb6ea05665fb393731ebd74b93325e254e77
+size 442563297

special_tokens_map.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "special_tokens_map_file": null, "name_or_path": "Maltehb/danish-bert-botxo", "do_basic_tokenize": true, "never_split": null}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec4c661f53a945e42c1dc4229d640ebe9d1c36b27566854d419a6100054d3623
+size 2351