| .. |
|
adapter_commons
|
99680b2d23
feat: soft prompts (#589)
|
пре 5 месеци |
|
attention
|
fa15bad2ea
chore: minor AMD fixes
|
пре 5 месеци |
|
common
|
ba371fbbbd
feat: AWQ marlin kernels (#603)
|
пре 5 месеци |
|
distributed
|
f91991f584
fix: f-string fixes
|
пре 5 месеци |
|
endpoints
|
f91991f584
fix: f-string fixes
|
пре 5 месеци |
|
engine
|
a4cbcfe59f
feat: disable logprob serialization to CPU for spec decode
|
пре 5 месеци |
|
executor
|
45a004874c
chore: allow specifying custom Executor
|
пре 5 месеци |
|
inputs
|
4f7d212b70
feat: remove vision language config
|
пре 5 месеци |
|
kv_quant
|
e42a78381a
feat: switch from pylint to ruff (#322)
|
пре 10 месеци |
|
lora
|
f91991f584
fix: f-string fixes
|
пре 5 месеци |
|
modeling
|
709628a74d
fix
|
пре 5 месеци |
|
multimodal
|
f91991f584
fix: f-string fixes
|
пре 5 месеци |
|
platforms
|
a3e26391e4
chore: add a wrapper for torch.inference_mode decorator
|
пре 5 месеци |
|
processing
|
6ac658b0d6
some small performance improvements
|
пре 5 месеци |
|
prompt_adapter
|
f91991f584
fix: f-string fixes
|
пре 5 месеци |
|
quantization
|
ba371fbbbd
feat: AWQ marlin kernels (#603)
|
пре 5 месеци |
|
spec_decode
|
a4cbcfe59f
feat: disable logprob serialization to CPU for spec decode
|
пре 5 месеци |
|
task_handler
|
f91991f584
fix: f-string fixes
|
пре 5 месеци |
|
transformers_utils
|
45a004874c
chore: allow specifying custom Executor
|
пре 5 месеци |
|
triton_utils
|
c8d398a4ae
feat: add custom triton cache manager
|
пре 5 месеци |
|
__init__.py
|
0c17c2a8a7
chore: add commit hash, clean up engine logs
|
пре 5 месеци |
|
_custom_ops.py
|
815736fc54
feat: add cuda kernels for sampling
|
пре 5 месеци |
|
_ipex_ops.py
|
9d7beaa5b9
chore: separate kv_scale into k_scale and v_scale
|
пре 5 месеци |
|
py.typed
|
1c988a48b2
fix logging and add py.typed
|
пре 1 година |
|
version.py
|
9038dea2df
fix: short commit hash import error
|
пре 5 месеци |
AlpinDale
709628a74d
fix