Commited backup

GeekTG/FTG-Modules/text_generator.py

@@ -0,0 +1,74 @@
+# By @vreply @pernel_kanic @nim1love @db0mb3r and geyporn by @tshipenchko
+
+import requests
+
+from .. import loader, utils
+
+
+def register(cb):
+    cb(TextGeneratorMod())
+
+
+class TextGeneratorMod(loader.Module):
+    "Generating text using machine learning"
+
+    strings = {
+        "name": "TextGenerator",
+        "no_text": "<strong>Empty message</strong>",
+        "wait": "<strong>Generating text...</strong>",
+    }
+
+    async def pfcmd(self, message):
+        """Generates text with Porfirevich: porfirevich.ru"""
+        text = utils.get_args_raw(message)
+        reply = await message.get_reply_message()
+
+        if text:
+            if reply:
+                text = reply.raw_text + text
+        elif reply:
+            text = reply.raw_text
+        else:
+            return await utils.answer(message, self.strings("no_text", message))
+
+        message = await utils.answer(message, self.strings("wait", message))
+        response = (
+            await utils.run_sync(
+                requests.post,
+                "https://pelevin.gpt.dobro.ai/generate/",
+                json={"prompt": text, "length": 30},
+            )
+        ).json()
+
+        return await utils.answer(
+            message, f"<strong>{text}</strong>" + response["replies"][-1]
+        )
+
+    async def gptcmd(self, message):
+        """Generates text with ruGPT-3 XL: russiannlp.github.io/rugpt-demo/"""
+        text = utils.get_args_raw(message)
+        reply = await message.get_reply_message()
+
+        if text:
+            if reply:
+                text = reply.raw_text + text
+        elif reply:
+            text = reply.raw_text
+        else:
+            return await utils.answer(message, self.strings("no_text", message))
+
+        message = await utils.answer(message, self.strings("wait", message))
+        response = (
+            await utils.run_sync(
+                requests.post,
+                "https://api.aicloud.sbercloud.ru/public/v1/public_inference/gpt3/predict",
+                json={"text": text},
+            )
+        ).json()
+        return await utils.answer(
+            message,
+            "<strong>"
+            + text
+            + "</strong>"
+            + response["predictions"].split(text.split()[-1], maxsplit=1)[1],
+        )