| .. |
|
__init__.py
|
04b53d2db5
chore: add initializer files
|
преди 1 година |
|
cache_engine.py
|
5be90c3859
Mamba infrastrucuture support (#586)
|
преди 6 месеца |
|
cpu_model_runner.py
|
99680b2d23
feat: soft prompts (#589)
|
преди 6 месеца |
|
cpu_worker.py
|
99680b2d23
feat: soft prompts (#589)
|
преди 6 месеца |
|
embedding_model_runner.py
|
99680b2d23
feat: soft prompts (#589)
|
преди 6 месеца |
|
model_runner.py
|
99680b2d23
feat: soft prompts (#589)
|
преди 6 месеца |
|
model_runner_base.py
|
5be90c3859
Mamba infrastrucuture support (#586)
|
преди 6 месеца |
|
neuron_model_runner.py
|
4599c98f99
feat: dynamic image size support for VLMs
|
преди 6 месеца |
|
neuron_worker.py
|
ae04f57ec1
feat: Pipeline Parallel support (#581)
|
преди 7 месеца |
|
openvino_model_runner.py
|
4f7d212b70
feat: remove vision language config
|
преди 6 месеца |
|
openvino_worker.py
|
1ff6d4c3d7
feat: support pipeline parallel on indivisible GPU count (#587)
|
преди 6 месеца |
|
tpu_model_runner.py
|
1cb06835a0
fix: TPU multimodal kwargs and outlines installation in TPU docker
|
преди 6 месеца |
|
tpu_worker.py
|
1ff6d4c3d7
feat: support pipeline parallel on indivisible GPU count (#587)
|
преди 6 месеца |
|
worker.py
|
d9f4c36edd
feat: Medusa speculative decoding support (#590)
|
преди 6 месеца |
|
worker_base.py
|
ddb28a80a3
fix: bump torch for rocm, unify CUDA_VISIBLE_DEVICES for cuda and rocm
|
преди 6 месеца |
|
xpu_model_runner.py
|
99680b2d23
feat: soft prompts (#589)
|
преди 6 месеца |
|
xpu_worker.py
|
99680b2d23
feat: soft prompts (#589)
|
преди 6 месеца |
AlpinDale
ddb28a80a3
fix: bump torch for rocm, unify CUDA_VISIBLE_DEVICES for cuda and rocm