first commit

README.md

@@ -0,0 +1,66 @@
+---
+language: 
+- sv
+thumbnail: 
+tags:
+- squad
+- translation
+license: apache-2.0
+datasets:
+- 
+metrics:
+- squad_v2
+---
+
+# Swedish BERT Fine-tuned on SQuAD v2
+
+This model is a fine-tuning checkpoint of Swedish BERT on SQuAD v2.
+
+## Training data
+
+Fine-tuning was done based on the pre-trained model [KB/bert-base-swedish-cased](https://huggingface.co/KB/bert-base-swedish-cased).
+
+Training and dev datasets are our
+[Swedish translation of SQuAD v2](https://github.com/susumu2357/SQuAD_v2_sv). 
+
+[Here](https://huggingface.co/datasets/susumu2357/squad_v2_sv) is the HuggingFace Datasets.
+
+
+## Hyperparameters
+```
+batch_size = 16
+n_epochs = 2
+max_seq_len = 386
+learning_rate = 3e-5
+warmup_steps = 2900    # warmup_proportion = 0.2
+doc_stride=128
+max_query_length=64
+```
+
+## Eval results
+```
+'exact': 66.72642524202223
+'f1': 70.11149581003404
+'total': 11156
+'HasAns_exact': 55.574745730186144
+'HasAns_f1': 62.821693965983044
+'HasAns_total': 5211
+'NoAns_exact': 76.50126156433979
+'NoAns_f1': 76.50126156433979
+'NoAns_total': 5945
+```
+
+## Limitations and bias
+
+This model may contain biases due to mistranslations of the SQuAD dataset.
+
+## BibTeX entry and citation info
+
+```bibtex
+@misc{svSQuADbert,
+  author = {Susumu Okazawa},
+  title = {Swedish BERT Fine-tuned on Swedish SQuAD 2.0},
+  year = {2021},
+  howpublished = {\url{}},
+}
+```

config.json

@@ -0,0 +1,22 @@
+{
+  "_name_or_path": "/content/drive/My Drive/NLP project/BERT_epoch2_warmup",
+  "architectures": [
+    "BertForQuestionAnswering"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "output_past": true,
+  "pad_token_id": 0,
+  "type_vocab_size": 2,
+  "vocab_size": 50325
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e307929e12fa22596cf0c3cb9e7b2bf5f993ad3d87a09315fd0a53cdaaade887
+size 496493853

special_tokens_map.json

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

tf_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca1f1b278623984160da108a2a63aaf8e0652c864b0840898d3e4792092bbfad
+size 499040016

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "special_tokens_map_file": "/root/.cache/torch/transformers/b7d6038768d76bb445ef775c850600834c30d86ade9a018032440242f7c1fd31.275045728fbf41c11d3dae08b8742c054377e18d92cc7b72b6351152a99b64e4", "tokenizer_file": null, "name_or_path": "/content/drive/My Drive/NLP project/BERT_epoch1_warmup/checkpoint-3000"}