assistant

main.py

@@ -1,6 +1,10 @@
 import io
+import logging
 import os
 import json
+import re
+import sys
+import traceback
 
 import requests
 from openai import OpenAI
@@ -15,6 +19,8 @@ ssl = None
 app.openai_key = os.environ.get("OPENAI_KEY", "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")
 app.openai_client = OpenAI(api_key=app.openai_key)
 
+#logging.basicConfig(level=logging.DEBUG, filename='/jkt-disk-01/app/mms/chatgpt-apache/chatgpt.log', format='%(asctime)s %(message)s')
+
 app.chat_messages = [
     {"role": "system",
      "content": "Please respond professionally and in a friendly manner, using the same language as the original request."}
@@ -63,12 +69,10 @@ def recommend(headlines, category):
                                     """
         }
         chat_messages.append(json_payload)
-        print(chat_messages)
         json_response = app.openai_client.chat.completions.create(model="gpt-3.5-turbo-1106",
                                                                   messages=chat_messages,
                                                                   response_format={"type": "json_object"}
                                                                   )
-        print(json_response.choices[0].message.content)
         return json.loads(json_response.choices[0].message.content)
     except Exception as error_print:
         app.logger.error(error_print)
@@ -96,7 +100,6 @@ def vision(message, image_url=None, image_b64=None):
             ],
         }
         chat_messages.append(json_payload)
-        print(chat_messages)
         json_response = app.openai_client.chat.completions.create(
             model="gpt-4o",
             messages=chat_messages,
@@ -110,7 +113,8 @@ def vision(message, image_url=None, image_b64=None):
 
 @app.route('/gpt', methods=['POST'])
 def gpt():
-    chat_messages = app.chat_messages.copy()
+    assistant_id = ""
+    chat_messages = []
     chat_model = "gpt-3.5-turbo"
     use_video = False
     suggest = False
@@ -130,12 +134,14 @@ def gpt():
         has_named_params = 'payload' in json_payload
         if 'payload' in json_payload:
             if 'predict_q' in json_payload:
-                predict_q = 5 if json_payload['predict_q'] > 4 else 0 if json_payload['predict_q'] < 1 else json_payload['predict_q']
+                predict_q = 5 if json_payload['predict_q'] > 4 else 0 if json_payload['predict_q'] < 1 else \
+                json_payload['predict_q']
             if 'num_choices' in json_payload:
-                num_choices = 5 if json_payload['num_choices'] > 4 else 1 if json_payload['num_choices'] < 2 else json_payload['num_choices']
+                num_choices = 5 if json_payload['num_choices'] > 4 else 1 if json_payload['num_choices'] < 2 else \
+                json_payload['num_choices']
             if 'use_video' in json_payload:
                 use_video = json_payload['use_video'] == "1"
-            if 'chat_model' in json_payload:
+            if 'chat_model' in json_payload and 'assistant_id' not in json_payload:
                 chat_model = json_payload['chat_model']
                 max_resp_token = 2048
             if 'translate' in json_payload:
@@ -160,6 +166,8 @@ def gpt():
                     chat_messages = app.chat_messages.copy()
                 json_payload['payload'][-1]['content'] = f"Please summarize this article:\n" + \
                                                          json_payload['payload'][-1]['content']
+            elif 'assistant_id' in json_payload:
+                assistant_id = json_payload['assistant_id']
             else:
                 chat_messages = app.chat_messages.copy()
             json_payload = json_payload['payload']
@@ -221,56 +229,70 @@ def gpt():
     result = {}
     try:
         n = num_choices
-        json_response = app.openai_client.chat.completions.create(model=chat_model,
-                                                                  messages=chat_messages,
-                                                                  max_tokens=max_resp_token, temperature=0.7, n=n)
-        app.logger.info(json_response.choices[0].message)
-        if has_named_params:
-            if suggest:
-                choices = json_response.choices
-                messages = [i.message for i in choices]
-                json_formatted = []
-                for message in messages:
-                    json_formatted.append({"role": "assistant", "content": message.content})
-                result = {"url": "", "message": json_formatted}
-            else:
-                if use_video:
-                    # TODO: to be implemented
-                    result = {"url": url_for('download_file', name="test.mp4", _external=True),
-                              "message": {"role": "assistant", "content": json_response.choices[0].message.content}}
-                else:
-                    result = {"role": "assistant", "content": json_response.choices[0].message.content}
-                if predict_q:
-                    query_q = {
-                        "role": "user",
-                        "content": f"Berikan {predict_q} pertanyaan lain yang akan saya ajukan berdasarkan percakapan kali ini dalam bentuk json array"
-                    }
-                    chat_messages.append(result)
-                    chat_messages.append(query_q)
-                    json_response_q = app.openai_client.chat.completions.create(model=chat_model,
-                                                                              messages=chat_messages,
-                                                                              max_tokens=max_resp_token,
-                                                                              temperature=0.2, response_format={"type": "json_object"})
-                    json_response_dict = json.loads(json_response_q.choices[0].message.content)
-                    print(json_response_dict)
-                    if json_response_dict is not None:
-                        if isinstance(json_response_dict, dict):
-                            if len(json_response_dict) > 1:
-                                qs = []
-                                for q in json_response_dict.values():
-                                    qs.append(q)
-                                json_response_dict = qs
-                            else:
-                                try:
-                                    first_key = next(iter(json_response_dict))
-                                    json_response_dict = json_response_dict[first_key]
-                                except StopIteration:
-                                    json_response_dict = []
-                        elif isinstance(json_response_dict, str):
-                            json_response_dict = [json_response_dict]
-                        result["predict_q"] = json_response_dict
+        if assistant_id and not suggest:
+            runs = app.openai_client.beta.threads.create_and_run_poll(
+                assistant_id=assistant_id,
+                thread={
+                    "messages": chat_messages
+                }
+            )
+            messages = list(app.openai_client.beta.threads.messages.list(thread_id=runs.thread_id, run_id=runs.id))
+            message_content = messages[0].content[0].text
+            app.logger.info(message_content.value)
+            pattern = re.compile(r"【\d+:\d+†\(?source\)?】")
+            filtered_message = pattern.sub("", message_content.value)
+            result = {"role": "assistant", "content": filtered_message}
         else:
-            result = {"role": "assistant", "content": json_response.choices[0].message.content}
+            json_response = app.openai_client.chat.completions.create(model=chat_model,
+                                                                      messages=chat_messages,
+                                                                      max_tokens=max_resp_token, temperature=0.7, n=n)
+            app.logger.info(json_response.choices[0].message)
+            if has_named_params:
+                if suggest:
+                    choices = json_response.choices
+                    messages = [i.message for i in choices]
+                    json_formatted = []
+                    for message in messages:
+                        json_formatted.append({"role": "assistant", "content": message.content})
+                    result = {"url": "", "message": json_formatted}
+                else:
+                    if use_video:
+                        # TODO: to be implemented
+                        result = {"url": url_for('download_file', name="test.mp4", _external=True),
+                                  "message": {"role": "assistant", "content": json_response.choices[0].message.content}}
+                    else:
+                        result = {"role": "assistant", "content": json_response.choices[0].message.content}
+            else:
+                result = {"role": "assistant", "content": json_response.choices[0].message.content}
+        if predict_q:
+            query_q = {
+                "role": "user",
+                "content": f"Berikan {predict_q} pertanyaan lain yang akan saya ajukan berdasarkan percakapan kali ini dalam bentuk json array"
+            }
+            chat_messages.append(result)
+            chat_messages.append(query_q)
+            json_response_q = app.openai_client.chat.completions.create(model=chat_model,
+                                                                        messages=chat_messages,
+                                                                        max_tokens=max_resp_token,
+                                                                        temperature=0.2,
+                                                                        response_format={"type": "json_object"})
+            json_response_dict = json.loads(json_response_q.choices[0].message.content)
+            if json_response_dict is not None:
+                if isinstance(json_response_dict, dict):
+                    if len(json_response_dict) > 1:
+                        qs = []
+                        for q in json_response_dict.values():
+                            qs.append(q)
+                        json_response_dict = qs
+                    else:
+                        try:
+                            first_key = next(iter(json_response_dict))
+                            json_response_dict = json_response_dict[first_key]
+                        except StopIteration:
+                            json_response_dict = []
+                elif isinstance(json_response_dict, str):
+                    json_response_dict = [json_response_dict]
+                result["predict_q"] = json_response_dict
     except Exception as error_print:
         app.logger.error(error_print)
         result = {}, 405
@@ -308,7 +330,8 @@ def train():
                 }
             )
             app.logger.info({"mock": "no", "status": job.status, "job_id": job.id})
-            return {"status": job.status, "job_id": job.id}
+            retval = {"status": job.status, "job_id": job.id}
+            return retval
         else:
             app.logger.info({"mock": "yes", "status": "ok"})
             return {"status": "ok"}
@@ -316,6 +339,7 @@ def train():
         app.logger.error({"status": "error", "message": "Training file not found"})
         return {"status": "error", "message": "Training file not found"}
 
+
 def train_with_id(job_id):
     try:
         job = app.openai_client.fine_tuning.jobs.retrieve(job_id)
@@ -329,6 +353,101 @@ def train_with_id(job_id):
         app.logger.error(error_print)
         return {"status": "Could not find job from id"}
 
+
+@app.route('/assistant/create', methods=['POST'])
+def assistant_create():
+    model_name = "gpt-3.5-turbo"
+    assistant_name = "Assistant"
+    assistant_ins = "Please respond professionally and in a friendly manner, using the same language as the original request."
+    assistant = None
+    if request.is_json:
+        request_form = request.json
+    else:
+        request_form = request.form.copy()
+    assistant_name = request_form.pop('name', assistant_name)
+    assistant_ins = request_form.pop('instructions', assistant_ins)
+    model_name = request_form.pop('model_name', model_name)
+    try:
+        assistant = app.openai_client.beta.assistants.create(
+            name=assistant_name,
+            instructions=assistant_ins,
+            model=model_name,
+            tools=[{"type": "file_search"}],
+            **request_form
+        )
+        if 'attachment1' in request.files:
+            resp_att = assistant_att()
+            if resp_att['status'] == 'completed':
+                resp_upd = assistant_update(assistant.id, resp_att['vector_store_id'])
+                assistant_updated = "1" if resp_upd['status'] == 'ok' else "0"
+            else:
+                assistant_updated = "0"
+            return {"status": "ok", "assistant_id": assistant.id, "assistant_updated": assistant_updated}
+        else:
+            return {"status": "ok", "assistant_id": assistant.id, "assistant_updated": "0"}
+    except ValueError as e:
+        return {"status": "error",
+                "message": "Failed to create assistant, please check whether your parameters are correct"}
+    except Exception:
+        return {"status": "error", "message": "Failed to create assistant, please try again"}
+
+
+@app.route('/assistant/attachment', methods=['POST'])
+def assistant_att():
+    vector_store_id = request.form.get('vector_store_id', '')
+    file_batch_id = request.form.get('file_batch_id', '')
+    attachments: list[str] = []
+    try:
+        if not vector_store_id or not file_batch_id:
+            if 'attachment1' not in request.files:
+                return {"status": "error", "message": "No file for attachments"}
+            else:
+                has_attachments = True
+                n = 1
+                while has_attachments:
+                    if f'attachment{n}' in request.files:
+                        retf = app.openai_client.files.create(
+                            file=(request.files[f'attachment{n}'].filename,
+                                  request.files[f'attachment{n}'].read()),
+                            purpose="assistants"
+                        )
+                        retf.filename = request.files[f'attachment{n}'].filename
+                        attachments.append(retf.id)
+                        n = n + 1
+                    else:
+                        has_attachments = False
+            vector_store = app.openai_client.beta.vector_stores.create()
+            file_batch = app.openai_client.beta.vector_stores.file_batches.create_and_poll(
+                vector_store_id=vector_store.id,
+                file_ids=attachments
+            )
+            return {"status": file_batch.status, "vector_store_id": vector_store.id, "file_batch_id": file_batch.id}
+        else:
+            file_batch = app.openai_client.beta.vector_stores.file_batches.retrieve(file_batch_id, vector_store_id=vector_store_id)
+            return {"status": file_batch.status}
+    except Exception as e:
+        app.logger.exception("error")
+        return {"status": "error", "message": "Upload attachment failed, please try again"}
+
+
+@app.route('/assistant/update', methods=['POST'])
+def assistant_update(aid=None, vid=None):
+    try:
+        if aid is not None and vid is not None:
+            assistant_id = aid
+            vector_store_id = vid
+        else:
+            assistant_id = request.form['assistant_id']
+            vector_store_id = request.form['vector_store_id']
+        app.openai_client.beta.assistants.update(
+            assistant_id=assistant_id,
+            tool_resources={"file_search": {"vector_store_ids": [vector_store_id]}},
+        )
+        return {"status": "ok"}
+    except Exception as e:
+        app.logger.exception("error")
+        return {"status": "error", "message": "Update assistant failed, please try again"}
+
 @app.route('/llama', methods=['POST'])
 def llama():
     max_char_msg = 500
@@ -375,7 +494,6 @@ def llama():
     return result
 
 
-
 # Press the green button in the gutter to run the script.
 if __name__ == '__main__':
     app.run(host='0.0.0.0', port=8348, debug=True, ssl_context=ssl)

requirements.txt

@@ -1,48 +1,7 @@
-aiohttp==3.8.4
-aiosignal==1.3.1
-annotated-types==0.6.0
-anyio==4.3.0
-async-timeout==4.0.2
-attrs==23.1.0
-blinker==1.6.2
-certifi==2023.5.7
-chardet==3.0.4
-charset-normalizer==3.1.0
-click==8.1.3
-distro==1.9.0
-exceptiongroup==1.2.0
-Flask==2.3.2
-frozenlist==1.3.3
-googletrans==3.0.0
-gunicorn==20.1.0
-h11==0.14.0
-h2==3.2.0
-hpack==3.0.0
-hstspreload==2023.1.1
-httpcore==1.0.5
-httpx==0.27.0
-hyperframe==5.2.0
-idna==2.10
-itsdangerous==2.1.2
-Jinja2==3.1.2
-jsonlines==4.0.0
-MarkupSafe==2.1.3
-mod-wsgi==4.9.4
-multidict==6.0.4
-numpy==1.26.4
-openai==1.16.0
-pandas==2.2.1
-pydantic==2.6.4
-pydantic_core==2.16.3
-python-dateutil==2.9.0.post0
-pytz==2024.1
-requests==2.31.0
-rfc3986==1.5.0
-six==1.16.0
-sniffio==1.3.0
-tqdm==4.65.0
-typing_extensions==4.10.0
-tzdata==2024.1
-urllib3==2.0.3
-Werkzeug==2.3.5
-yarl==1.9.2
+requests~=2.32.3
+openai~=1.35.10
+Flask~=3.0.3
+pandas~=2.2.2
+jsonlines~=4.0.0
+Werkzeug~=3.0.3
+mod-wsgi==4.9.4