Unverified Commit 9cc733b3 authored by Jerry Zhang's avatar Jerry Zhang Committed by GitHub
Browse files

move apply_torchao_config_ to model_runner (#2342)

parent d693ec04
...@@ -7,13 +7,15 @@ from typing import Dict, Set ...@@ -7,13 +7,15 @@ from typing import Dict, Set
import torch import torch
def torchao_quantize_param_data(param: torch.Tensor, torchao_config: str): def apply_torchao_config_to_model_(
"""Quantize a Tensor with torchao quantization specified by torchao_config model: torch.nn.Module, torchao_config: str, filter_fn=None
):
"""Quantize a modelwith torchao quantization specified by torchao_config
Args: Args:
`param`: weight parameter of the linear module `model`: a model to be quantized based on torchao_config
`torchao_config`: type of quantization and their arguments we want to use to `torchao_config` (str): type of quantization and their arguments we want to use to
quantize the Tensor, e.g. int4wo-128 means int4 weight only quantization with group_size quantize the model, e.g. int4wo-128 means int4 weight only quantization with group_size
128 128
""" """
# Lazy import to suppress some warnings # Lazy import to suppress some warnings
...@@ -26,12 +28,12 @@ def torchao_quantize_param_data(param: torch.Tensor, torchao_config: str): ...@@ -26,12 +28,12 @@ def torchao_quantize_param_data(param: torch.Tensor, torchao_config: str):
) )
from torchao.quantization.observer import PerRow, PerTensor from torchao.quantization.observer import PerRow, PerTensor
dummy_linear = torch.nn.Linear(param.shape[1], param.shape[0], bias=False) if torchao_config == "" or torchao_config is None:
dummy_linear.weight = param return model
if "int8wo" in torchao_config: elif "int8wo" in torchao_config:
quantize_(dummy_linear, int8_weight_only()) quantize_(model, int8_weight_only(), filter_fn=filter_fn)
elif "int8dq" in torchao_config: elif "int8dq" in torchao_config:
quantize_(dummy_linear, int8_dynamic_activation_int8_weight()) quantize_(model, int8_dynamic_activation_int8_weight(), filter_fn=filter_fn)
elif "int4wo" in torchao_config: elif "int4wo" in torchao_config:
group_size = int(torchao_config.split("-")[-1]) group_size = int(torchao_config.split("-")[-1])
assert group_size in [ assert group_size in [
...@@ -40,13 +42,13 @@ def torchao_quantize_param_data(param: torch.Tensor, torchao_config: str): ...@@ -40,13 +42,13 @@ def torchao_quantize_param_data(param: torch.Tensor, torchao_config: str):
128, 128,
256, 256,
], f"int4wo groupsize needs to be one of [32, 64, 128, 256] but got {group_size}" ], f"int4wo groupsize needs to be one of [32, 64, 128, 256] but got {group_size}"
quantize_(dummy_linear, int4_weight_only(group_size=group_size)) quantize_(model, int4_weight_only(group_size=group_size), filter_fn=filter_fn)
elif "fp8wo" in torchao_config: elif "fp8wo" in torchao_config:
from torchao.quantization import float8_weight_only from torchao.quantization import float8_weight_only
# this requires newer hardware # this requires newer hardware
# [rank0]: AssertionError: fp8e4nv data type is not supported on CUDA arch < 89 # [rank0]: AssertionError: fp8e4nv data type is not supported on CUDA arch < 89
quantize_(dummy_linear, float8_weight_only()) quantize_(model, float8_weight_only(), filter_fn=filter_fn)
elif "fp8dq" in torchao_config: elif "fp8dq" in torchao_config:
granularity = torchao_config.split("-")[-1] granularity = torchao_config.split("-")[-1]
GRANULARITY_MAP = { GRANULARITY_MAP = {
...@@ -57,39 +59,13 @@ def torchao_quantize_param_data(param: torch.Tensor, torchao_config: str): ...@@ -57,39 +59,13 @@ def torchao_quantize_param_data(param: torch.Tensor, torchao_config: str):
granularity in GRANULARITY_MAP granularity in GRANULARITY_MAP
), f"Supported granularity are: {GRANULARITY_MAP.keys()}, got {granularity}" ), f"Supported granularity are: {GRANULARITY_MAP.keys()}, got {granularity}"
quantize_( quantize_(
dummy_linear, model,
float8_dynamic_activation_float8_weight( float8_dynamic_activation_float8_weight(
granularity=GRANULARITY_MAP[granularity] granularity=GRANULARITY_MAP[granularity]
), ),
filter_fn=filter_fn,
) )
else: else:
raise ValueError(f"Unexpected config: {torchao_config}") raise ValueError(f"Unexpected config: {torchao_config}")
return dummy_linear.weight return model
def apply_torchao_config_(
self: torch.nn.Module,
params_dict: Dict[str, torch.Tensor],
param_suffixes: Set[str],
) -> None:
"""A util function used for quantizing the weight parameters after they are loaded if
self.torchao_config is specified
Args:
`self`: the model we want to quantize
`params_dict`: dictionary mapping from param_name to the parameter Tensor
`param_suffixes`: a set of suffixes, we'll quantize the Tensor matching these suffixes
Returns:
None, the `params_dict` is modified inplace and the weights of `self` model are quantized
"""
if self.torchao_config:
for param_suffix in param_suffixes:
for name in params_dict:
param = params_dict[name]
if param_suffix in name and param.ndim == 2:
params_dict[name] = torchao_quantize_param_data(
param, self.torchao_config
)
self.load_state_dict(params_dict, assign=True)
...@@ -38,6 +38,7 @@ from sglang.srt.layers.attention.torch_native_backend import TorchNativeAttnBack ...@@ -38,6 +38,7 @@ from sglang.srt.layers.attention.torch_native_backend import TorchNativeAttnBack
from sglang.srt.layers.attention.triton_backend import TritonAttnBackend from sglang.srt.layers.attention.triton_backend import TritonAttnBackend
from sglang.srt.layers.logits_processor import LogitsProcessorOutput from sglang.srt.layers.logits_processor import LogitsProcessorOutput
from sglang.srt.layers.sampler import Sampler from sglang.srt.layers.sampler import Sampler
from sglang.srt.layers.torchao_utils import apply_torchao_config_to_model_
from sglang.srt.lora.lora_manager import LoRAManager from sglang.srt.lora.lora_manager import LoRAManager
from sglang.srt.managers.schedule_batch import global_server_args_dict from sglang.srt.managers.schedule_batch import global_server_args_dict
from sglang.srt.mem_cache.memory_pool import ( from sglang.srt.mem_cache.memory_pool import (
...@@ -159,6 +160,13 @@ class ModelRunner: ...@@ -159,6 +160,13 @@ class ModelRunner:
else: else:
self.torch_tp_applied = False self.torch_tp_applied = False
def filter_fn(module, fqn):
return "proj" in fqn
apply_torchao_config_to_model_(
self.model, global_server_args_dict["torchao_config"], filter_fn
)
# Init memory pool and attention backends # Init memory pool and attention backends
if server_args.lora_paths is not None: if server_args.lora_paths is not None:
self.init_lora_manager() self.init_lora_manager()
......
...@@ -35,12 +35,10 @@ from sglang.srt.layers.linear import ( ...@@ -35,12 +35,10 @@ from sglang.srt.layers.linear import (
from sglang.srt.layers.logits_processor import LogitsProcessor from sglang.srt.layers.logits_processor import LogitsProcessor
from sglang.srt.layers.quantization.base_config import QuantizationConfig from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.radix_attention import RadixAttention from sglang.srt.layers.radix_attention import RadixAttention
from sglang.srt.layers.torchao_utils import apply_torchao_config_
from sglang.srt.layers.vocab_parallel_embedding import ( from sglang.srt.layers.vocab_parallel_embedding import (
ParallelLMHead, ParallelLMHead,
VocabParallelEmbedding, VocabParallelEmbedding,
) )
from sglang.srt.managers.schedule_batch import global_server_args_dict
from sglang.srt.model_executor.forward_batch_info import ForwardBatch from sglang.srt.model_executor.forward_batch_info import ForwardBatch
from sglang.srt.model_loader.loader import DefaultModelLoader from sglang.srt.model_loader.loader import DefaultModelLoader
from sglang.srt.model_loader.weight_utils import default_weight_loader from sglang.srt.model_loader.weight_utils import default_weight_loader
...@@ -290,7 +288,6 @@ class Grok1ForCausalLM(nn.Module): ...@@ -290,7 +288,6 @@ class Grok1ForCausalLM(nn.Module):
super().__init__() super().__init__()
self.config = config self.config = config
self.quant_config = quant_config self.quant_config = quant_config
self.torchao_config = global_server_args_dict["torchao_config"]
self.model = Grok1Model(config, quant_config=quant_config) self.model = Grok1Model(config, quant_config=quant_config)
self.lm_head = ParallelLMHead(config.vocab_size, config.hidden_size) self.lm_head = ParallelLMHead(config.vocab_size, config.hidden_size)
self.logits_processor = LogitsProcessor(config) self.logits_processor = LogitsProcessor(config)
...@@ -374,8 +371,6 @@ class Grok1ForCausalLM(nn.Module): ...@@ -374,8 +371,6 @@ class Grok1ForCausalLM(nn.Module):
) )
weight_loader(param, loaded_weight) weight_loader(param, loaded_weight)
apply_torchao_config_(self, params_dict, set(["proj.weight"]))
class Grok1ModelForCausalLM(Grok1ForCausalLM): class Grok1ModelForCausalLM(Grok1ForCausalLM):
"""An alias for backward-compatbility.""" """An alias for backward-compatbility."""
......
...@@ -36,12 +36,10 @@ from sglang.srt.layers.logits_processor import LogitsProcessor, LogitsProcessorO ...@@ -36,12 +36,10 @@ from sglang.srt.layers.logits_processor import LogitsProcessor, LogitsProcessorO
from sglang.srt.layers.pooler import Pooler, PoolingType from sglang.srt.layers.pooler import Pooler, PoolingType
from sglang.srt.layers.quantization.base_config import QuantizationConfig from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.radix_attention import RadixAttention from sglang.srt.layers.radix_attention import RadixAttention
from sglang.srt.layers.torchao_utils import apply_torchao_config_
from sglang.srt.layers.vocab_parallel_embedding import ( from sglang.srt.layers.vocab_parallel_embedding import (
ParallelLMHead, ParallelLMHead,
VocabParallelEmbedding, VocabParallelEmbedding,
) )
from sglang.srt.managers.schedule_batch import global_server_args_dict
from sglang.srt.model_executor.forward_batch_info import ForwardBatch from sglang.srt.model_executor.forward_batch_info import ForwardBatch
from sglang.srt.model_loader.weight_utils import default_weight_loader from sglang.srt.model_loader.weight_utils import default_weight_loader
from sglang.srt.utils import make_layers from sglang.srt.utils import make_layers
...@@ -304,7 +302,6 @@ class LlamaForCausalLM(nn.Module): ...@@ -304,7 +302,6 @@ class LlamaForCausalLM(nn.Module):
super().__init__() super().__init__()
self.config = config self.config = config
self.quant_config = quant_config self.quant_config = quant_config
self.torchao_config = global_server_args_dict["torchao_config"]
self.model = LlamaModel(config, quant_config=quant_config) self.model = LlamaModel(config, quant_config=quant_config)
# Llama 3.2 1B Insturct set tie_word_embeddings to True # Llama 3.2 1B Insturct set tie_word_embeddings to True
# Llama 3.1 8B Insturct set tie_word_embeddings to False # Llama 3.1 8B Insturct set tie_word_embeddings to False
...@@ -424,8 +421,6 @@ class LlamaForCausalLM(nn.Module): ...@@ -424,8 +421,6 @@ class LlamaForCausalLM(nn.Module):
weight_loader = getattr(param, "weight_loader", default_weight_loader) weight_loader = getattr(param, "weight_loader", default_weight_loader)
weight_loader(param, loaded_weight) weight_loader(param, loaded_weight)
apply_torchao_config_(self, params_dict, set(["proj.weight"]))
def get_weights_by_name( def get_weights_by_name(
self, name: str, truncate_size: int = 100, tp_size: int = 1 self, name: str, truncate_size: int = 100, tp_size: int = 1
) -> Optional[torch.Tensor]: ) -> Optional[torch.Tensor]:
......
...@@ -34,12 +34,10 @@ from sglang.srt.layers.linear import ( ...@@ -34,12 +34,10 @@ from sglang.srt.layers.linear import (
from sglang.srt.layers.logits_processor import LogitsProcessor from sglang.srt.layers.logits_processor import LogitsProcessor
from sglang.srt.layers.quantization.base_config import QuantizationConfig from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.radix_attention import RadixAttention from sglang.srt.layers.radix_attention import RadixAttention
from sglang.srt.layers.torchao_utils import apply_torchao_config_
from sglang.srt.layers.vocab_parallel_embedding import ( from sglang.srt.layers.vocab_parallel_embedding import (
ParallelLMHead, ParallelLMHead,
VocabParallelEmbedding, VocabParallelEmbedding,
) )
from sglang.srt.managers.schedule_batch import global_server_args_dict
from sglang.srt.model_executor.forward_batch_info import ForwardBatch from sglang.srt.model_executor.forward_batch_info import ForwardBatch
from sglang.srt.model_loader.weight_utils import default_weight_loader from sglang.srt.model_loader.weight_utils import default_weight_loader
...@@ -295,7 +293,6 @@ class MixtralForCausalLM(nn.Module): ...@@ -295,7 +293,6 @@ class MixtralForCausalLM(nn.Module):
super().__init__() super().__init__()
self.config = config self.config = config
self.quant_config = quant_config self.quant_config = quant_config
self.torchao_config = global_server_args_dict["torchao_config"]
self.model = MixtralModel(config, quant_config=quant_config, prefix="model") self.model = MixtralModel(config, quant_config=quant_config, prefix="model")
self.lm_head = ParallelLMHead(config.vocab_size, config.hidden_size) self.lm_head = ParallelLMHead(config.vocab_size, config.hidden_size)
self.logits_processor = LogitsProcessor(config) self.logits_processor = LogitsProcessor(config)
...@@ -387,7 +384,5 @@ class MixtralForCausalLM(nn.Module): ...@@ -387,7 +384,5 @@ class MixtralForCausalLM(nn.Module):
) )
weight_loader(param, loaded_weight) weight_loader(param, loaded_weight)
apply_torchao_config_(self, params_dict, set(["proj.weight"]))
EntryClass = MixtralForCausalLM EntryClass = MixtralForCausalLM
...@@ -17,13 +17,11 @@ from sglang.srt.layers.logits_processor import LogitsProcessor, LogitsProcessorO ...@@ -17,13 +17,11 @@ from sglang.srt.layers.logits_processor import LogitsProcessor, LogitsProcessorO
from sglang.srt.layers.pooler import Pooler, PoolingType from sglang.srt.layers.pooler import Pooler, PoolingType
from sglang.srt.layers.quantization.base_config import QuantizationConfig from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.radix_attention import RadixAttention from sglang.srt.layers.radix_attention import RadixAttention
from sglang.srt.layers.torchao_utils import apply_torchao_config_
from sglang.srt.layers.vocab_parallel_embedding import ( from sglang.srt.layers.vocab_parallel_embedding import (
DEFAULT_VOCAB_PADDING_SIZE, DEFAULT_VOCAB_PADDING_SIZE,
ParallelLMHead, ParallelLMHead,
VocabParallelEmbedding, VocabParallelEmbedding,
) )
from sglang.srt.managers.schedule_batch import global_server_args_dict
from sglang.srt.model_executor.forward_batch_info import ForwardBatch from sglang.srt.model_executor.forward_batch_info import ForwardBatch
from sglang.srt.model_loader.weight_utils import default_weight_loader from sglang.srt.model_loader.weight_utils import default_weight_loader
from sglang.srt.utils import make_layers from sglang.srt.utils import make_layers
...@@ -348,7 +346,6 @@ class Phi3SmallForCausalLM(nn.Module): ...@@ -348,7 +346,6 @@ class Phi3SmallForCausalLM(nn.Module):
quant_config=quant_config, quant_config=quant_config,
prefix="model", prefix="model",
) )
self.torchao_config = global_server_args_dict["torchao_config"]
self.vocab_size = config.vocab_size self.vocab_size = config.vocab_size
self.mup_width_multiplier = config.mup_width_multiplier self.mup_width_multiplier = config.mup_width_multiplier
self.lm_head = ParallelLMHead( self.lm_head = ParallelLMHead(
...@@ -441,7 +438,5 @@ class Phi3SmallForCausalLM(nn.Module): ...@@ -441,7 +438,5 @@ class Phi3SmallForCausalLM(nn.Module):
weight_loader = getattr(param, "weight_loader", default_weight_loader) weight_loader = getattr(param, "weight_loader", default_weight_loader)
weight_loader(param, loaded_weight) weight_loader(param, loaded_weight)
apply_torchao_config_(self, params_dict, set(["proj.weight"]))
EntryClass = Phi3SmallForCausalLM EntryClass = Phi3SmallForCausalLM
...@@ -40,12 +40,10 @@ from sglang.srt.layers.linear import ( ...@@ -40,12 +40,10 @@ from sglang.srt.layers.linear import (
from sglang.srt.layers.logits_processor import LogitsProcessor from sglang.srt.layers.logits_processor import LogitsProcessor
from sglang.srt.layers.quantization.base_config import QuantizationConfig from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.radix_attention import RadixAttention from sglang.srt.layers.radix_attention import RadixAttention
from sglang.srt.layers.torchao_utils import apply_torchao_config_
from sglang.srt.layers.vocab_parallel_embedding import ( from sglang.srt.layers.vocab_parallel_embedding import (
ParallelLMHead, ParallelLMHead,
VocabParallelEmbedding, VocabParallelEmbedding,
) )
from sglang.srt.managers.schedule_batch import global_server_args_dict
from sglang.srt.model_executor.forward_batch_info import ForwardBatch from sglang.srt.model_executor.forward_batch_info import ForwardBatch
from sglang.srt.model_loader.weight_utils import default_weight_loader from sglang.srt.model_loader.weight_utils import default_weight_loader
...@@ -352,7 +350,6 @@ class Qwen2MoeForCausalLM(nn.Module): ...@@ -352,7 +350,6 @@ class Qwen2MoeForCausalLM(nn.Module):
super().__init__() super().__init__()
self.config = config self.config = config
self.quant_config = quant_config self.quant_config = quant_config
self.torchao_config = global_server_args_dict["torchao_config"]
self.model = Qwen2MoeModel(config, quant_config) self.model = Qwen2MoeModel(config, quant_config)
self.lm_head = ParallelLMHead( self.lm_head = ParallelLMHead(
config.vocab_size, config.hidden_size, quant_config=quant_config config.vocab_size, config.hidden_size, quant_config=quant_config
...@@ -445,7 +442,5 @@ class Qwen2MoeForCausalLM(nn.Module): ...@@ -445,7 +442,5 @@ class Qwen2MoeForCausalLM(nn.Module):
) )
weight_loader(param, loaded_weight) weight_loader(param, loaded_weight)
apply_torchao_config_(self, params_dict, set(["proj.weight"]))
EntryClass = Qwen2MoeForCausalLM EntryClass = Qwen2MoeForCausalLM
...@@ -58,12 +58,10 @@ from sglang.srt.layers.layernorm import RMSNorm ...@@ -58,12 +58,10 @@ from sglang.srt.layers.layernorm import RMSNorm
from sglang.srt.layers.logits_processor import LogitsProcessor, LogitsProcessorOutput from sglang.srt.layers.logits_processor import LogitsProcessor, LogitsProcessorOutput
from sglang.srt.layers.quantization.base_config import QuantizationConfig from sglang.srt.layers.quantization.base_config import QuantizationConfig
from sglang.srt.layers.radix_attention import RadixAttention from sglang.srt.layers.radix_attention import RadixAttention
from sglang.srt.layers.torchao_utils import apply_torchao_config_
from sglang.srt.layers.vocab_parallel_embedding import ( from sglang.srt.layers.vocab_parallel_embedding import (
ParallelLMHead, ParallelLMHead,
VocabParallelEmbedding, VocabParallelEmbedding,
) )
from sglang.srt.managers.schedule_batch import global_server_args_dict
from sglang.srt.model_executor.forward_batch_info import ForwardBatch from sglang.srt.model_executor.forward_batch_info import ForwardBatch
from sglang.srt.model_loader.weight_utils import default_weight_loader from sglang.srt.model_loader.weight_utils import default_weight_loader
...@@ -392,7 +390,6 @@ class TorchNativeLlamaForCausalLM(nn.Module): ...@@ -392,7 +390,6 @@ class TorchNativeLlamaForCausalLM(nn.Module):
super().__init__() super().__init__()
self.config = config self.config = config
self.quant_config = quant_config self.quant_config = quant_config
self.torchao_config = global_server_args_dict["torchao_config"]
self.supports_torch_tp = True self.supports_torch_tp = True
self.model = LlamaModel(config, quant_config=quant_config) self.model = LlamaModel(config, quant_config=quant_config)
if self.config.tie_word_embeddings: if self.config.tie_word_embeddings:
...@@ -503,8 +500,6 @@ class TorchNativeLlamaForCausalLM(nn.Module): ...@@ -503,8 +500,6 @@ class TorchNativeLlamaForCausalLM(nn.Module):
weight_loader = getattr(param, "weight_loader", default_weight_loader) weight_loader = getattr(param, "weight_loader", default_weight_loader)
weight_loader(param, loaded_weight) weight_loader(param, loaded_weight)
apply_torchao_config_(self, params_dict, set(["proj.weight"]))
class TorchNativePhi3ForCausalLM(TorchNativeLlamaForCausalLM): class TorchNativePhi3ForCausalLM(TorchNativeLlamaForCausalLM):
pass pass
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment