 AlpinDale
|
5289c14b24
feat: Asymmetric Tensor Parallel (#594)
|
преди 4 месеца |
|
AlpinDale
|
99680b2d23
feat: soft prompts (#589)
|
преди 4 месеца |
 AlpinDale
|
4f7d212b70
feat: remove vision language config
|
преди 4 месеца |
|
AlpinDale
|
4599c98f99
feat: dynamic image size support for VLMs
|
преди 4 месеца |
|
AlpinDale
|
5be90c3859
Mamba infrastrucuture support (#586)
|
преди 4 месеца |
|
AlpinDale
|
ae04f57ec1
feat: Pipeline Parallel support (#581)
|
преди 4 месеца |
|
AlpinDale
|
cdff8e89f9
feat: introduce `DraftModelRunner`
|
преди 5 месеца |
|
AlpinDale
|
405bb74612
Control plane comms refactor (#573)
|
преди 5 месеца |
|
AlpinDale
|
8d77c69cbd
feat: support image processor and add llava example
|
преди 5 месеца |
|
AlpinDale
|
e4ea3da1ad
fix: tensor parallel with embedding model
|
преди 5 месеца |
|
AlpinDale
|
de62ceb18c
refactor: eliminate parallel worker per-step task scheduling overhead
|
преди 5 месеца |
|
AlpinDale
|
a94de94c44
refactor: combine the prefill and decode into a single API (#553)
|
преди 5 месеца |
|
AlpinDale
|
50b7c13db0
refactor: attention selector (#552)
|
преди 5 месеца |
|
AlpinDale
|
be8154a8a0
feat: proper embeddings API with e5-mistral-7b support
|
преди 5 месеца |
