Skip to main content

Fastspeed

Fastspeed is a toolkit for serving LLM.

Requirements

  • Python >= 3.6

Installation

pip install fastspeed

Example

  • server
fastspeed --model model_name_or_path  # default model_hub=modelscope
fastspeed --model model_name_or_path --model_hub hf
  • client
curl -X 'POST' \
  'http://127.0.0.1:8000/chat/completions' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "messages": [
    {
      "role": "user",
      "content": "Hello"
    }
  ],
  "model": ""
}'

Change Log

0.1.1 (2024-09-15)

  • Suport half, cuda and trust_remote_code=True

0.1.0 (2024-08-29)

  • Fastapi with single GPU, model from huggingface or modelscope

Release files for fastspeed 0.1.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for fastspeed 0.1.1
File Size Uploaded
fastspeed-0.1.1.tar.gz 4.5 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for fastspeed 0.1.1
File Interpreter ABI Platform
fastspeed-0.1.1-py2.py3-none-any.whl Python 2, Python 3 none any Details

Total release size: 8.9 kB

Release files / fastspeed-0.1.1.tar.gz

Download URL fastspeed-0.1.1.tar.gz
Size 4.5 kB
Tags Source
SHA-256 checksum
How to use checksums
5d9d82b8d6159f0e5bd6d424c422b6156ce8c630298fecc6515bd8c920d06516
BLAKE2b-256 checksum
How to use checksums
dc0e25e591b93593288add097758fbf478bae7a79eefcbb28f216f1b7998c925
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/5.1.1 CPython/3.11.10

Release files / fastspeed-0.1.1-py2.py3-none-any.whl

Download URL fastspeed-0.1.1-py2.py3-none-any.whl
Size 4.4 kB
Tags Python 2 Python 3
SHA-256 checksum
How to use checksums
0be6cac2258b54a85a9595a1688279543707834095f5fdef0466e54c006a4464
BLAKE2b-256 checksum
How to use checksums
68512b9cd08f5fdcdf0320d018e98466c40edb0e303e5ac838694123e0654e47
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/5.1.1 CPython/3.11.10

Release history Release notifications | RSS feed

This release

0.1.1 This release

2 release files

0.1.0

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page