[ Kernel ] Fp8 Channelwise Weight Support (#6487)

vllm/config.py

@@ -238,7 +238,8 @@ class ModelConfig:
                     f"{self.quantization} quantization is currently not "
                     f"supported in ROCm.")
             if (self.quantization
-                    not in ("fp8", "marlin", "gptq_marlin_24", "gptq_marlin")):
+                    not in ("fp8", "marlin", "gptq_marlin_24", "gptq_marlin",
+                            "compressed_tensors")):
                 logger.warning(
                     "%s quantization is not fully "
                     "optimized yet. The speed can be slower than "

vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py

@@ -13,7 +13,8 @@ from vllm.model_executor.layers.quantization.compressed_tensors.schemes import (
     CompressedTensorsWNA16)
 from vllm.model_executor.layers.quantization.compressed_tensors.utils import (
     CompressionFormat, QuantizationArgs, QuantizationStrategy,
-    QuantizationType, find_first_name_or_class_match)
+    QuantizationType, find_first_name_or_class_match,
+    is_activation_quantization_format)
 from vllm.platforms import current_platform
 
 
@@ -132,10 +133,11 @@ class CompressedTensorsConfig(QuantizationConfig):
         # Confirm weight scheme is supported.
         is_symmetric_weight = weight_quant.symmetric
         is_static_weight = not weight_quant.dynamic
-        is_per_tensor_weight = (
+        is_per_tensor_or_channel_weight = (weight_quant.strategy in [
-            weight_quant.strategy == QuantizationStrategy.TENSOR)
+            QuantizationStrategy.TENSOR, QuantizationStrategy.CHANNEL
+        ])
         if not (is_symmetric_weight and is_static_weight
-                and is_per_tensor_weight):
+                and is_per_tensor_or_channel_weight):
             return False
 
         # Dynamic quantization is always supported if weights supported.
@@ -167,6 +169,7 @@ class CompressedTensorsConfig(QuantizationConfig):
     def _get_schema(self, weight_quant: BaseModel,
                     input_quant: BaseModel) -> "CompressedTensorsScheme":
 
+        # Detect If Mixed Precision
         if self._is_wNa16_group_channel(weight_quant, input_quant):
             self._check_gptq_and_marlin_can_run()
             if (self.quant_format == CompressionFormat.marlin_24.value
@@ -182,11 +185,12 @@ class CompressedTensorsConfig(QuantizationConfig):
                     strategy=weight_quant.strategy,
                     group_size=weight_quant.group_size)
 
-        if (self.quant_format == CompressionFormat.int_quantized.value or
+        # Detect If Activation Quantization.
-                self.quant_format == CompressionFormat.float_quantized.value):
+        if is_activation_quantization_format(self.quant_format):
             if self._is_fp8_w8a8(weight_quant, input_quant):
                 return CompressedTensorsW8A8Fp8(
-                    input_dynamic=input_quant.dynamic)
+                    strategy=weight_quant.strategy,
+                    is_static_input_scheme=(not input_quant.dynamic))
 
             if self._is_static_tensor_w8a8(weight_quant, input_quant):
                 return CompressedTensorsW8A8Int8(

vllm/model_executor/layers/quantization/compressed_tensors/schemes/compressed_tensors_w8a8_fp8.py

@@ -1,11 +1,15 @@
 from typing import Callable, List, Optional
 
 import torch
+from torch.nn import Parameter
 
 from vllm.model_executor.layers.quantization.compressed_tensors.schemes import (
     CompressedTensorsScheme)
+from vllm.model_executor.layers.quantization.compressed_tensors.utils import (
+    QuantizationStrategy)
 from vllm.model_executor.layers.quantization.utils.w8a8_utils import (
-    apply_fp8_linear, create_per_tensor_scale_param, cutlass_fp8_supported,
+    apply_fp8_linear, create_per_channel_scale_param,
+    create_per_tensor_scale_param, cutlass_fp8_supported,
     requantize_with_max_scale)
 from vllm.model_executor.utils import set_weight_attrs
 
@@ -14,39 +18,56 @@ __all__ = ["CompressedTensorsW8A8Fp8"]
 
 class CompressedTensorsW8A8Fp8(CompressedTensorsScheme):
 
-    def __init__(self, input_dynamic: bool):
+    def __init__(self, strategy: str, is_static_input_scheme: bool):
-        self.input_dynamic = input_dynamic
+        self.strategy = strategy
+        self.is_static_input_scheme = is_static_input_scheme
         self.cutlass_fp8_supported = cutlass_fp8_supported()
 
-    # W8A8-Fp8 kernels support only per-tensor and per-channel cases.
+        # On Lovelace, fail for now if channelwise.
-    # So if we have a fused module (QKV, MLP) with per tensor scales (thus N
+        # TODO: (@tms) fallback
-    # scales being passed to the kernel), we requantize with a single scale.
+        if (not self.cutlass_fp8_supported
+                and self.strategy == QuantizationStrategy.CHANNEL):
+            raise ValueError(
+                "Channelwise fp8 quantization requires vLLM's custom "
+                "cutlass kernels, which are not supported on your device."
+                "Consider quantizing with per tensor scales or upgrading "
+                "to Hopper.")
+
     def process_weights_after_loading(self, layer) -> None:
-        # Dequant -> Quant with max scale.
+        # If per tensor, when we have a fused module (e.g. QKV) with per
+        # tensor scales (thus N scales being passed to the kernel),
+        # requantize so we can always run per tensor
+        if self.strategy == QuantizationStrategy.TENSOR:
             max_w_scale, weight = requantize_with_max_scale(
                 weight=layer.weight,
                 weight_scale=layer.weight_scale,
                 logical_widths=layer.logical_widths,
             )
 
-        # Update layer with new values.
+            layer.weight = Parameter(weight.t(), requires_grad=False)
-        layer.weight = torch.nn.Parameter(weight.t(), requires_grad=False)
+            layer.weight_scale = Parameter(max_w_scale, requires_grad=False)
-        layer.weight_scale = torch.nn.Parameter(max_w_scale,
+
-                                                requires_grad=False)
+        # If channelwise, scales are already lined up, so just transpose.
-        if self.input_dynamic:
+        elif self.strategy == QuantizationStrategy.CHANNEL:
-            layer.input_scale = None
+            assert self.cutlass_fp8_supported
+            weight = layer.weight
+            layer.weight = Parameter(weight.t(), requires_grad=False)
+
         else:
-            layer.input_scale = torch.nn.Parameter(layer.input_scale.max(),
+            raise ValueError(f"Unknown quantization strategy {self.strategy}")
+
+        # INPUT SCALE
+        if self.is_static_input_scheme:
+            layer.input_scale = Parameter(layer.input_scale.max(),
                                           requires_grad=False)
+        else:
+            layer.input_scale = None
 
     def create_weights(self, layer: torch.nn.Module,
                        output_partition_sizes: List[int],
                        input_size_per_partition: int,
                        params_dtype: torch.dtype, weight_loader: Callable,
                        **kwargs):
-
-        del params_dtype
-
         output_size_per_partition = sum(output_partition_sizes)
         layer.logical_widths = output_partition_sizes
 
@@ -63,12 +84,17 @@ class CompressedTensorsW8A8Fp8(CompressedTensorsScheme):
         })
 
         # WEIGHT SCALE
+        if self.strategy == QuantizationStrategy.CHANNEL:
+            weight_scale = create_per_channel_scale_param(
+                output_partition_sizes, weight_loader=weight_loader)
+        else:
+            assert self.strategy == QuantizationStrategy.TENSOR
             weight_scale = create_per_tensor_scale_param(
                 output_partition_sizes, weight_loader=weight_loader)
         layer.register_parameter("weight_scale", weight_scale)
 
         # INPUT SCALE
-        if not self.input_dynamic:
+        if self.is_static_input_scheme:
             input_scale = create_per_tensor_scale_param(
                 output_partition_sizes, weight_loader=weight_loader)
             layer.register_parameter("input_scale", input_scale)

vllm/model_executor/layers/quantization/compressed_tensors/utils.py

@@ -9,6 +9,7 @@ from torch.nn import Module
 class CompressionFormat(Enum):
     dense = "dense"
     sparse_bitmask = "sparse-bitmask"
+    naive_quantized = "naive-quantized"
     float_quantized = "float-quantized"
     int_quantized = "int-quantized"
     pack_quantized = "pack-quantized"
@@ -76,6 +77,15 @@ class QuantizationArgs(BaseModel):
     )
 
 
+def is_activation_quantization_format(format: str) -> bool:
+    _ACTIVATION_QUANTIZATION_FORMATS = [
+        CompressionFormat.naive_quantized.value,
+        CompressionFormat.int_quantized.value,
+        CompressionFormat.float_quantized.value
+    ]
+    return format in _ACTIVATION_QUANTIZATION_FORMATS
+
+
 def find_first_name_or_class_match(
         name: str,
         module: Module,