Skip to content
Snippets Groups Projects

Add streaming support for LLM and TTS models for faster response

1 unresolved thread

Commit also adds unit test basis as logic is getting more complex

Merge request reports

Loading
Loading

Activity

Filter activity
  • Approvals
  • Assignees & reviewers
  • Comments (from bots)
  • Comments (from users)
  • Commits & branches
  • Edits
  • Labels
  • Lock status
  • Mentions
  • Merge request status
  • Tracking
6 from http.server import BaseHTTPRequestHandler,HTTPServer
7 from os import curdir, sep
8 import torch
9 from diffusers import AutoPipelineForText2Image, LCMScheduler
10 import simplejson
11 import base64
12 from io import BytesIO
13 from PIL import Image
14
15
16
17 PORT_NUMBER = 8005
18
19 #This class will handles any incoming request from
20 #the browser
21 class myHandler(BaseHTTPRequestHandler):
  • Jari Helaakoski added 10 commits

    added 10 commits

    • 25c21ee3 - Structured output model
    • 144573f2 - Merge branch 'structured-output-model' into 'main'
    • 757925c1 - Prototype streaming
    • eb8338e1 - Ignore build directory
    • 06962f21 - Fixes for streaming proto
    • 9ecfbfdc - Add unit test + many fixes
    • af4c1f77 - Fixes for tts streaming
    • 6d18bf45 - More tests and code fixes. Improve the test app functionality.
    • 2341a9c9 - Move servers into separate directory
    • c4006fa8 - Fixes

    Compare with previous version

  • Please register or sign in to reply
    Loading