Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update LLM microservice output to binary #123

Merged
merged 1 commit into from
May 31, 2024
Merged

Update LLM microservice output to binary #123

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 3 additions & 20 deletions comps/llms/text-generation/tgi/llm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,18 +21,6 @@
from comps import GeneratedDoc, LLMParamsDoc, ServiceType, opea_microservices, register_microservice


@traceable(run_type="tool")
def post_process_text(text: str):
if text == " ":
return "data: @#$\n\n"
if text == "\n":
return "data: <br/>\n\n"
if text.isspace():
return None
new_text = text.replace("Answer: ", "").replace("Human: ", "").replace(" ", "@#$")
return f"data: {new_text}\n\n"


@register_microservice(
name="opea_service@llm_tgi",
service_type=ServiceType.LLM,
Expand Down Expand Up @@ -61,14 +49,9 @@ async def stream_generator():
chat_response = ""
async for text in llm.astream(input.query):
chat_response += text
processed_text = post_process_text(text)
if text and processed_text:
if "</s>" in text:
res = text.split("</s>")[0]
if res != "":
yield res
break
yield processed_text
chunk_repr = repr(text.encode("utf-8"))
print(f"[llm - chat_stream] chunk:{chunk_repr}")
yield f"data: {chunk_repr}\n\n"
print(f"[llm - chat_stream] stream response: {chat_response}")
yield "data: [DONE]\n\n"

Expand Down
Loading