No project description provided

Navigation

Verified details

These details have been verified by PyPI
Maintainers
Avatar for mvkvc from gravatar.com mvkvc

Unverified details

These details have not been verified by PyPI
Project links
Meta
  • License: MIT License (MIT License)
  • Author: Marko Vukovic
  • Tags nbdev, jupyter, notebook, python
  • Requires: Python >=3.10
Classifiers

Project description

wwd

wwd: westworld but available on PyPI

Components

  • api: Unified API for LM-related inference
  • nbs: nbdev notebooks for main package
  • notes: Research notes
  • papers: Research papers
  • ra: Research assistant tool
  • sh: Shell scripts
  • vendor: Git submodules
  • wwd: Main package

Goals

  • Data pipelines
  • Benchmarks
  • Create intent based API for LM-related inference
  • Create RAG+ research assistant tool
  • Reimplement
    • Embedding
    • Reranking
    • Transformer (NanoGPT/Mistral/LLaMA)
    • Multi-modal transformer (LLaVA)
    • RNN (RWKV)
    • SSM (Mamba)
    • Spiking NN
    • Forward-forward
  • Research additions
    • RAG over previous interactions
    • Fused backprop step at inference for transformers
  • Research new approaches
    • Forward-training architectures

Project details

Verified details

These details have been verified by PyPI
Maintainers
Avatar for mvkvc from gravatar.com mvkvc

Unverified details

These details have not been verified by PyPI
Project links
Meta
  • License: MIT License (MIT License)
  • Author: Marko Vukovic
  • Tags nbdev, jupyter, notebook, python
  • Requires: Python >=3.10
Classifiers

Release history Release notifications | RSS feed

This version

0.0.1

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

wwd-0.0.1.tar.gz (4.5 kB view details)

Uploaded Source

Built Distribution

wwd-0.0.1-py3-none-any.whl (3.7 kB view details)

Uploaded Python 3

File details

Details for the file wwd-0.0.1.tar.gz.

File metadata

  • Download URL: wwd-0.0.1.tar.gz
  • Upload date:
  • Size: 4.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.0.0 CPython/3.10.12

File hashes

Hashes for wwd-0.0.1.tar.gz
Algorithm Hash digest
SHA256 f67791c0daed5a46e45ab7130bbddd5f0605ab324396649bff26db01a4229358
MD5 cd7a47d1072e4eecf75b3ae4f2ff0c53
BLAKE2b-256 d890e6d05730f2b6b2a21e7fffb753496a01016f2081792d8e38e9513a863ee1

See more details on using hashes here.

File details

Details for the file wwd-0.0.1-py3-none-any.whl.

File metadata

  • Download URL: wwd-0.0.1-py3-none-any.whl
  • Upload date:
  • Size: 3.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.0.0 CPython/3.10.12

File hashes

Hashes for wwd-0.0.1-py3-none-any.whl
Algorithm Hash digest
SHA256 750885c3972fd4d127c089aa9511c947338f49e2b4aa08c893442edc8235a730
MD5 762268e9e99958aeb4b8ef0dc59d54f2
BLAKE2b-256 0787e9d2db3de5cd8ac5f595672c3cd4e5220ba7dff00d610c4eb746ab748744

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page