Smart Summary is now working!

There's no preferences dialog, so you can't really adjust the prompt
or the model it uses. The default settings work well for me. You may
want to tweak them depending on your model preferences and compute
budget. (Not many can afford to run Llama3-8B at high
quantization. Conversely, you might have a better GPU than me and wish
to run a 27B model or bigger.)

2 files changed, 109 insertions, 3 deletions

diff --git a/po/bowl.pot b/po/bowl.pot
index 24e87b5..92da214 100644
--- a/po/bowl.pot
+++ b/po/bowl.pot
@@ -8,7 +8,7 @@ msgid ""
 msgstr ""
 "Project-Id-Version: bowl\n"
 "Report-Msgid-Bugs-To: \n"
-"POT-Creation-Date: 2024-09-01 18:02+0300\n"
+"POT-Creation-Date: 2024-09-04 15:59+0300\n"
 "PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
 "Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
 "Language-Team: LANGUAGE <LL@li.org>\n"
@@ -130,3 +130,52 @@ msgstr ""
 #: src/lib.rs:331
 msgid "Micropub access token for {}"
 msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:6
+msgid "LLM API endpoint"
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:7
+msgid "Ollama API endpoint used to query an LLM for Smart Summary."
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:13
+msgid "Smart Summary LLM"
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:25
+msgid "LLM system prompt"
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:26
+msgid ""
+"The system prompt provided to the LLM. For best results, it should instruct "
+"the LLM to provide a one-sentence summary of the document it receives. The "
+"default system prompt is tested for Llama 3.1-8B and should work for posts "
+"written mainly in English. Performance with other languages is untested."
+msgstr ""
+
+#. TRANSLATORS: please keep the link intact
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:15
+msgid ""
+"The model that Ollama will load to produce summaries. Available models can "
+"be seen at <a href=\"https://ollama.com/library\">Ollama library</a>."
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:38
+msgid "Smart Summary prompt prefix"
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:39
+msgid ""
+"What the text is prefixed with when pasted into the LLM prompt. Something "
+"like \"Summarize this text:\" works well."
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:47
+msgid "Smart Summary prompt suffix"
+msgstr ""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:48
+msgid "Append this to the prompt after the article text."
+msgstr ""
diff --git a/po/ru.po b/po/ru.po
index 29af65e..8ce26e5 100644
--- a/po/ru.po
+++ b/po/ru.po
@@ -29,7 +29,7 @@ msgstr "Минималистичная Micropub-утилита для напис
 #. Translators: Search terms to find this application. Do NOT translate or localize the semicolons! The list MUST also end with a semicolon!
 #: data/xyz.vikanezrimaya.kittybox.Bowl.desktop.in.in:10
 msgid "Micropub;IndieWeb;Kittybox;"
-msgstr "Miropub;IndieWeb;Kittybox"
+msgstr "Micropub;IndieWeb;Kittybox;"
 
 #. TRANSLATORS: please keep the newline and `<b>` tags
 #: src/components/smart_summary.rs:47
@@ -58,7 +58,7 @@ msgstr "Текст"
 
 #: src/components/post_editor.rs:280
 msgid "Visibility"
-msgstr ""
+msgstr "Видимость"
 
 #: src/components/post_editor.rs:493
 msgid "Smart Summary error: {}"
@@ -135,3 +135,60 @@ msgstr "Опубликовать"
 #: src/lib.rs:331
 msgid "Micropub access token for {}"
 msgstr "Токен доступа Micropub для {}"
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:6
+msgid "LLM API endpoint"
+msgstr "Точка API LLM"
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:7
+msgid "Ollama API endpoint used to query an LLM for Smart Summary."
+msgstr "API Ollama, которое используется, чтобы сгенерировать Умную Выжимку с помощью языковой модели."
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:13
+msgid "Smart Summary LLM"
+msgstr "Модель для Умной Выжимки"
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:22
+msgid "LLM system prompt"
+msgstr "Системная вводная для LLM"
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:23
+msgid ""
+"The system prompt provided to the LLM. For best results, it should instruct "
+"the LLM to provide a one-sentence summary of the document it receives. The "
+"default system prompt is tested for Llama 3.1-8B and should work for posts "
+"written mainly in English. Performance with other languages is untested."
+msgstr ""
+"Системная вводная, которая будет передана языковой модели. Для достижения "
+"наилучших результатов она должна содержать в себе указание для модели — "
+"описать суть документа одним предложением. Вводная, указанная по умолчанию, "
+"протестирована для Llama 3.1-8B и лучше всего работает со статьями на "
+"английском. Результаты для других языков не гарантированы."
+
+#. TRANSLATORS: please keep the link intact
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:15
+msgid ""
+"The model that Ollama will load to produce summaries. Available models can "
+"be seen at <a href=\"https://ollama.com/library\">Ollama library</a>."
+msgstr ""
+"Языковая модель, которую Ollama использует для извлечения содержания текста."
+"Доступные модели можно увидеть в <a href=\"https://ollama.com/library\">библиотеке Ollama</a>."
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:38
+msgid "Smart Summary prompt prefix"
+msgstr "Префикс вводной для Умной Выжимки"
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:39
+msgid ""
+"What the text is prefixed with when pasted into the LLM prompt. Something "
+"like \"Summarize this text:\" works well."
+msgstr "Что приписывается к началу текста для вводной языковой модели. Пример: "
+"\"Опиши смысл этого текста одним предложением:\""
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:47
+msgid "Smart Summary prompt suffix"
+msgstr "Суффикс вводной Умной Выжимки"
+
+#: data/xyz.vikanezrimaya.kittybox.Bowl.gschema.xml.in:48
+msgid "Append this to the prompt after the article text."
+msgstr "Что приписывается к вводной после текста статьи."