5 Commits

Author SHA1 Message Date
Ryan Chen
7161c09a4e do not fully delete lol 2025-10-24 08:47:59 -04:00
Ryan Chen
68d73b62e8 Instituting LLM fallback to OpenAI if gaming PC is not on 2025-10-24 08:44:08 -04:00
Ryan Chen
6b616137d3 oops 2025-10-23 22:45:14 -04:00
Ryan Chen
841b6ebd4f i commit 2025-10-23 22:39:26 -04:00
Ryan Chen
45a5e92aee Added conversation history (#4)
Reviewed-on: #4
Co-authored-by: Ryan Chen <ryan@torrtle.co>
Co-committed-by: Ryan Chen <ryan@torrtle.co>
2025-10-23 22:29:12 -04:00
4 changed files with 36 additions and 29 deletions

View File

@@ -23,6 +23,7 @@ RUN uv pip install --system -e .
# Copy application code
COPY *.py ./
COPY blueprints ./blueprints
COPY startup.sh ./
RUN chmod +x startup.sh
@@ -43,4 +44,4 @@ ENV PYTHONPATH=/app
ENV CHROMADB_PATH=/app/chromadb
# Run the startup script
CMD ["./startup.sh"]
CMD ["./startup.sh"]

1
blueprints/__init__.py Normal file
View File

@@ -0,0 +1 @@
# Blueprints package

57
llm.py
View File

@@ -4,9 +4,14 @@ from ollama import Client
from openai import OpenAI
import logging
from dotenv import load_dotenv
load_dotenv()
logging.basicConfig(level=logging.INFO)
TRY_OLLAMA = os.getenv("TRY_OLLAMA", False)
class LLMClient:
def __init__(self):
@@ -30,31 +35,35 @@ class LLMClient:
prompt: str,
system_prompt: str,
):
# Instituting a fallback if my gaming PC is not on
if self.PROVIDER == "ollama":
response = self.ollama_client.chat(
model="gemma3:4b",
messages=[
{
"role": "system",
"content": system_prompt,
},
{"role": "user", "content": prompt},
],
)
print(response)
output = response.message.content
elif self.PROVIDER == "openai":
response = self.openai_client.responses.create(
model="gpt-4o-mini",
input=[
{
"role": "system",
"content": system_prompt,
},
{"role": "user", "content": prompt},
],
)
output = response.output_text
try:
response = self.ollama_client.chat(
model="gemma3:4b",
messages=[
{
"role": "system",
"content": system_prompt,
},
{"role": "user", "content": prompt},
],
)
output = response.message.content
return output
except Exception as e:
logging.error(f"Could not connect to OLLAMA: {str(e)}")
response = self.openai_client.responses.create(
model="gpt-4o-mini",
input=[
{
"role": "system",
"content": system_prompt,
},
{"role": "user", "content": prompt},
],
)
output = response.output_text
return output

View File

@@ -197,10 +197,6 @@ def filter_indexed_files(docs):
if __name__ == "__main__":
args = parser.parse_args()
if args.reindex:
with sqlite3.connect("./visited.db") as conn:
c = conn.cursor()
c.execute("DELETE FROM indexed_documents")
logging.info("Fetching documents from Paperless-NGX")
ppngx = PaperlessNGXService()
docs = ppngx.get_data()