 AlpinDale
|
d8f9f0ec16
fix: prefix prefill kernels for fp32 data type
|
5 kuukautta sitten |
|
AlpinDale
|
cda0e93a10
abstract away the platform for device capability
|
5 kuukautta sitten |
|
AlpinDale
|
7d79c0e726
chore: use nvml query to avoid accidental cuda initialization
|
5 kuukautta sitten |
|
AlpinDale
|
f40b809d3b
allow using v2 block manager with sliding window
|
6 kuukautta sitten |
|
AlpinDale
|
c3065aefaf
non-power-of-2 for prefix prefill with alibi
|
6 kuukautta sitten |
|
AlpinDale
|
aa15d3dc0f
sliding window in prefix prefill kernel
|
6 kuukautta sitten |
 AlpinDale
|
fca911ee0a
vLLM Upstream Sync (#526)
|
7 kuukautta sitten |
|
AlpinDale
|
66b7bc4415
sliding window in prefix kernel
|
8 kuukautta sitten |
|
AlpinDale
|
9d81716bfd
[v0.5.3] Release Candidate (#388)
|
8 kuukautta sitten |
