Skip to content

Commit

Permalink
[Misc] Add ignored layers for fp8 quantization (vllm-project#6657)
Browse files Browse the repository at this point in the history
  • Loading branch information
mgoin authored and fialhocoelho committed Jul 24, 2024
1 parent d379e0a commit e49b9d6
Show file tree
Hide file tree
Showing 4 changed files with 57 additions and 47 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,9 @@
from pydantic import BaseModel, Field
from torch.nn import Module

from vllm.model_executor.layers.quantization.utils.quant_utils import (
FUSED_LAYER_NAME_MAPPING)


class CompressionFormat(Enum):
dense = "dense"
Expand Down Expand Up @@ -86,13 +89,6 @@ def is_activation_quantization_format(format: str) -> bool:
return format in _ACTIVATION_QUANTIZATION_FORMATS


# fused_name: List[shard_name]
_FUSED_LAYER_NAME_MAPPING = {
"qkv_proj": ["q_proj", "k_proj", "v_proj"],
"gate_up_proj": ["gate_proj", "up_proj"]
}


def should_ignore_layer(layer_name: Optional[str],
ignore: Iterable[str]) -> bool:
if layer_name is None:
Expand All @@ -106,8 +102,8 @@ def should_ignore_layer(layer_name: Optional[str],
# in the safetensors checkpoint. So, we convert the name
# from the fused version to unfused + check to make sure that
# each shard of the fused layer has the same scheme.
if proj_name in _FUSED_LAYER_NAME_MAPPING:
shard_proj_names = _FUSED_LAYER_NAME_MAPPING[proj_name]
if proj_name in FUSED_LAYER_NAME_MAPPING:
shard_proj_names = FUSED_LAYER_NAME_MAPPING[proj_name]

# Convert fused_name --> [shard_names]
shard_names = [
Expand Down
39 changes: 3 additions & 36 deletions vllm/model_executor/layers/quantization/fbgemm_fp8.py
Original file line number Diff line number Diff line change
Expand Up @@ -11,21 +11,15 @@
QuantizationConfig, QuantizeMethodBase)
from vllm.model_executor.layers.quantization.utils.marlin_utils_fp8 import (
apply_fp8_marlin_linear, prepare_fp8_layer_for_marlin)
from vllm.model_executor.layers.quantization.utils.quant_utils import (
is_layer_skipped)
from vllm.model_executor.layers.quantization.utils.w8a8_utils import (
apply_fp8_linear, create_per_channel_scale_param)
from vllm.model_executor.utils import set_weight_attrs
from vllm.platforms import current_platform

logger = init_logger(__name__)

# Note: this is a hack. We should update each model to register the
# stacked params and get it from there instead in a future PR.
# fused_name: List[shard_name]
_FUSED_LAYER_NAME_MAPPING = {
"qkv_proj": ["q_proj", "k_proj", "v_proj"],
"gate_up_proj": ["gate_proj", "up_proj"]
}


class FBGEMMFp8Config(QuantizationConfig):
"""Config class for FBGEMM Fp8."""
Expand Down Expand Up @@ -62,37 +56,10 @@ def from_config(cls, config: Dict[str, Any]) -> "FBGEMMFp8Config":
input_scale_ub = cls.get_from_keys(config, ["activation_scale_ub"])
return cls(ignore_list=ignore_list, input_scale_ub=input_scale_ub)

def _is_layer_skipped(self, prefix: str) -> bool:
# prefix: model.layers.0.self_attn.q_proj
# proj_name: q_proj
proj_name = prefix.split(".")[-1]
if proj_name in _FUSED_LAYER_NAME_MAPPING:
shard_prefixes = [
prefix.replace(proj_name, shard_proj_name)
for shard_proj_name in _FUSED_LAYER_NAME_MAPPING[proj_name]
]

is_skipped = None
for shard_prefix in shard_prefixes:
is_shard_skipped = shard_prefix in self.ignore_list

if is_skipped is None:
is_skipped = is_shard_skipped
elif is_shard_skipped != is_skipped:
raise ValueError(
f"Detected some but not all shards of {prefix} "
"are quantized. All shards of fused layers "
"to have the same precision.")
else:
is_skipped = prefix in self.ignore_list

assert is_skipped is not None
return is_skipped

def get_quant_method(self, layer: torch.nn.Module,
prefix: str) -> Optional["QuantizeMethodBase"]:
if isinstance(layer, LinearBase):
if self._is_layer_skipped(prefix):
if is_layer_skipped(prefix, self.ignore_list):
return UnquantizedLinearMethod()
return FBGEMMFp8LinearMethod(self)
return None
Expand Down
13 changes: 11 additions & 2 deletions vllm/model_executor/layers/quantization/fp8.py
Original file line number Diff line number Diff line change
Expand Up @@ -8,12 +8,15 @@
from vllm.logger import init_logger
from vllm.model_executor.layers.fused_moe import (FusedMoE, FusedMoEMethodBase,
fused_moe)
from vllm.model_executor.layers.linear import LinearBase, LinearMethodBase
from vllm.model_executor.layers.linear import (LinearBase, LinearMethodBase,
UnquantizedLinearMethod)
from vllm.model_executor.layers.quantization.base_config import (
QuantizationConfig, QuantizeMethodBase)
from vllm.model_executor.layers.quantization.kv_cache import BaseKVCacheMethod
from vllm.model_executor.layers.quantization.utils.marlin_utils_fp8 import (
apply_fp8_marlin_linear, prepare_fp8_layer_for_marlin)
from vllm.model_executor.layers.quantization.utils.quant_utils import (
is_layer_skipped)
from vllm.model_executor.layers.quantization.utils.w8a8_utils import (
all_close_1d, apply_fp8_linear, create_per_tensor_scale_param,
cutlass_fp8_supported, per_tensor_dequantize, requantize_with_max_scale)
Expand All @@ -33,6 +36,7 @@ def __init__(
self,
is_checkpoint_fp8_serialized: bool = False,
activation_scheme: str = "dynamic",
ignored_layers: Optional[List[str]] = None,
) -> None:
self.is_checkpoint_fp8_serialized = is_checkpoint_fp8_serialized
if is_checkpoint_fp8_serialized:
Expand All @@ -42,6 +46,7 @@ def __init__(
raise ValueError(
f"Unsupported activation scheme {activation_scheme}")
self.activation_scheme = activation_scheme
self.ignored_layers = ignored_layers or []

@classmethod
def get_name(cls) -> str:
Expand All @@ -64,14 +69,18 @@ def from_config(cls, config: Dict[str, Any]) -> "Fp8Config":
quant_method = cls.get_from_keys(config, ["quant_method"])
is_checkpoint_fp8_serialized = ("fp8" in quant_method)
activation_scheme = cls.get_from_keys(config, ["activation_scheme"])
ignored_layers = cls.get_from_keys_or(config, ["ignored_layers"], None)
return cls(is_checkpoint_fp8_serialized=is_checkpoint_fp8_serialized,
activation_scheme=activation_scheme)
activation_scheme=activation_scheme,
ignored_layers=ignored_layers)

def get_quant_method(self, layer: torch.nn.Module,
prefix: str) -> Optional["QuantizeMethodBase"]:
from vllm.attention.layer import Attention # Avoid circular import

if isinstance(layer, LinearBase):
if is_layer_skipped(prefix, self.ignored_layers):
return UnquantizedLinearMethod()
return Fp8LinearMethod(self)
elif isinstance(layer, FusedMoE):
return Fp8MoEMethod(self)
Expand Down
38 changes: 38 additions & 0 deletions vllm/model_executor/layers/quantization/utils/quant_utils.py
Original file line number Diff line number Diff line change
@@ -1,10 +1,48 @@
"""This file is used for /tests and /benchmarks"""
from typing import List

import numpy
import torch

SUPPORTED_NUM_BITS = [4, 8]
SUPPORTED_GROUP_SIZES = [-1, 32, 64, 128]

# Note: this is a hack. We should update each model to register the
# stacked params and get it from there instead in a future PR.
# fused_name: List[shard_name]
FUSED_LAYER_NAME_MAPPING = {
"qkv_proj": ["q_proj", "k_proj", "v_proj"],
"gate_up_proj": ["gate_proj", "up_proj"]
}


def is_layer_skipped(prefix: str, ignored_layers: List[str]) -> bool:
# prefix: model.layers.0.self_attn.q_proj
# proj_name: q_proj
proj_name = prefix.split(".")[-1]
if proj_name in FUSED_LAYER_NAME_MAPPING:
shard_prefixes = [
prefix.replace(proj_name, shard_proj_name)
for shard_proj_name in FUSED_LAYER_NAME_MAPPING[proj_name]
]

is_skipped = None
for shard_prefix in shard_prefixes:
is_shard_skipped = shard_prefix in ignored_layers

if is_skipped is None:
is_skipped = is_shard_skipped
elif is_shard_skipped != is_skipped:
raise ValueError(
f"Detected some but not all shards of {prefix} "
"are quantized. All shards of fused layers "
"to have the same precision.")
else:
is_skipped = prefix in ignored_layers

assert is_skipped is not None
return is_skipped


def get_pack_factor(num_bits):
assert num_bits in SUPPORTED_NUM_BITS, f"Unsupported num_bits = {num_bits}"
Expand Down

0 comments on commit e49b9d6

Please sign in to comment.