 AlpinDale
|
4599c98f99
feat: dynamic image size support for VLMs
|
7 месяцев назад |
 AlpinDale
|
5be90c3859
Mamba infrastrucuture support (#586)
|
7 месяцев назад |
|
AlpinDale
|
ae04f57ec1
feat: Pipeline Parallel support (#581)
|
7 месяцев назад |
|
AlpinDale
|
cdff8e89f9
feat: introduce `DraftModelRunner`
|
7 месяцев назад |
|
AlpinDale
|
c0c336aaa3
refactor: registry for processing model inputs; quick_gelu; clip model support
|
7 месяцев назад |
|
AlpinDale
|
426a13ab73
fix: pass multi_modal_kwargs to CPU model runner
|
7 месяцев назад |
|
AlpinDale
|
405bb74612
Control plane comms refactor (#573)
|
7 месяцев назад |
|
AlpinDale
|
fdabb55a4d
fix: wrong multi_modal_input format for CPU
|
7 месяцев назад |
|
AlpinDale
|
8d77c69cbd
feat: support image processor and add llava example
|
7 месяцев назад |
|
AlpinDale
|
f6250c5516
move dockerfiles to root; fix cpu build
|
8 месяцев назад |
|
AlpinDale
|
a94de94c44
refactor: combine the prefill and decode into a single API (#553)
|
8 месяцев назад |
|
AlpinDale
|
50b7c13db0
refactor: attention selector (#552)
|
8 месяцев назад |
|
AlpinDale
|
0e062e66d3
set block size at init
|
8 месяцев назад |
|
AlpinDale
|
35ae01d7ba
refactor: attention metadata term
|
8 месяцев назад |
|
AlpinDale
|
aed64884c6
feat: prompt logprobs with chunked prefill (#539)
|
8 месяцев назад |
|
AlpinDale
|
fca911ee0a
vLLM Upstream Sync (#526)
|
8 месяцев назад |
|
AlpinDale
|
9d81716bfd
[v0.5.3] Release Candidate (#388)
|
10 месяцев назад |
