| .. |
|
__init__.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
cache_engine.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
cpu_model_runner.py
|
65a59bbb6b
cpu: raise error if using encoder-decoder models (#1027)
|
пре 2 недеља |
|
cpu_worker.py
|
f2b6dc3872
cpu: add support for W8A8 quantization via compressed-tensor (#1017)
|
пре 2 недеља |
|
embedding_model_runner.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
enc_dec_model_runner.py
|
a985143768
core: add cuda graph support for encoder-decoder models (#1051)
|
пре 2 недеља |
|
model_runner.py
|
a985143768
core: add cuda graph support for encoder-decoder models (#1051)
|
пре 2 недеља |
|
model_runner_base.py
|
304e1e5a8a
core: dump model runner inputs during crash (#1023)
|
пре 2 недеља |
|
multi_step_model_runner.py
|
1390915778
multi-step: add support for flashinfer attention backend (#1033)
|
пре 2 недеља |
|
multi_step_tpu_worker.py
|
4b1b658855
tpu: implement multi-step scheduling (#1046)
|
пре 2 недеља |
|
multi_step_worker.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
neuron_model_runner.py
|
145e554a4d
neuron: add 8bit quantization for Neuron (#994)
|
пре 2 недеља |
|
neuron_worker.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
openvino_model_runner.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
openvino_worker.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
tpu_model_runner.py
|
4b1b658855
tpu: implement multi-step scheduling (#1046)
|
пре 2 недеља |
|
tpu_worker.py
|
a50548c0b9
tpu: use XLA rank for persistent cache path (#989)
|
пре 2 недеља |
|
utils.py
|
a985143768
core: add cuda graph support for encoder-decoder models (#1051)
|
пре 2 недеља |
|
worker.py
|
a113309876
kernel: add meta functions for ops to prevent graph breaks (#1019)
|
пре 2 недеља |
|
worker_base.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
xpu_model_runner.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
|
xpu_worker.py
|
3bb0f07461
chore: rename `task_handler` to `worker` (#985)
|
пре 2 недеља |
AlpinDale
a985143768
core: add cuda graph support for encoder-decoder models (#1051)