Add dataset file

.gitattributes

@@ -53,3 +53,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.jpg filter=lfs diff=lfs merge=lfs -text
 *.jpeg filter=lfs diff=lfs merge=lfs -text
 *.webp filter=lfs diff=lfs merge=lfs -text
+ficbook_replaces.jsonl filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,40 @@
 ---
 license: apache-2.0
+task_categories:
+- text-generation
+language:
+- ru
+size_categories:
+- 1M<n<10M
 ---
+
+Texts inverse normalized obtained from [ficbook](https://huggingface.co/datasets/IlyaGusev/ficbook) dataset.
+Normalized using [these notebooks](https://github.com/saarus72/text_normalization) for a personal [russian normalization model](https://huggingface.co/saarus72/russian_text_normalizer) (avaliable on [HF Space](https://huggingface.co/spaces/saarus72/russian-text-normalization) as well).
+
+All put into single `jsonl` file with lines like (beautified):
+```json
+{
+  "replaces": [
+    {
+      "text_from": "Боль во всем теле...Боже...я так и знала.... ",
+      "text_to": "Боль во всем теле...Боже...я так и знала.... "
+    },
+    {
+      "text_from": "5",
+      "text_to": "Пятая"
+    },
+    {
+      "text_from": " точка буквально ныла от боли, ушиб будет тот еще...садиться не смогу дня ",
+      "text_to": " точка буквально ныла от боли, ушиб будет тот еще...садиться не смогу дня "
+    },
+    {
+      "text_from": "3",
+      "text_to": "три"
+    },
+    {
+      "text_from": " так, точно.",
+      "text_to": " так, точно."
+    }
+  ]
+}
+```

ficbook_replaces.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a310932c068b485ef2f95076049723366cdc7c54361138023164181819ffb66
+size 1911508020