Speed up docs build (#44635)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
@@ -87,8 +87,9 @@ class QuarkConfig(QuantizationConfig):
|
||||
Interface for models to update module names referenced in
|
||||
quantization configs in order to reflect the vllm model structure
|
||||
|
||||
:param hf_to_vllm_mapper: maps from hf model structure (the assumed
|
||||
structure of the qconfig) to vllm model structure
|
||||
Args:
|
||||
hf_to_vllm_mapper: maps from hf model structure (the assumed
|
||||
structure of the qconfig) to vllm model structure
|
||||
"""
|
||||
quant_config_with_hf_to_vllm_mapper: dict[str, Any] = {}
|
||||
|
||||
@@ -724,7 +725,9 @@ class QuarkKVCacheMethod(BaseKVCacheMethod):
|
||||
"""
|
||||
Validator for the kv cache configuration. Useful for controlling the
|
||||
kv cache quantization schemes, that are being supported in vLLM
|
||||
:param kv_cache_config: the quark kv cache scheme
|
||||
|
||||
Args:
|
||||
kv_cache_config: the quark kv cache scheme
|
||||
"""
|
||||
if kv_cache_config is None:
|
||||
return
|
||||
|
||||
@@ -38,11 +38,11 @@ class QuarkScheme(ABC):
|
||||
Run the forward pass for the particular scheme. This is where
|
||||
scheme-specific dequant/quant steps/kernels should be applied.
|
||||
|
||||
:param layer: torch.nn.Module with the registered weights and
|
||||
other parameters relevant to the particular scheme.
|
||||
:param x: input to the layer
|
||||
:param bias: bias parameter
|
||||
|
||||
Args:
|
||||
layer: torch.nn.Module with the registered weights and
|
||||
other parameters relevant to the particular scheme.
|
||||
x: input to the layer
|
||||
bias: bias parameter
|
||||
"""
|
||||
raise NotImplementedError
|
||||
|
||||
|
||||
Reference in New Issue
Block a user