[Misc] Update gptq_marlin to use new vLLMParameters (#7281)

.buildkite/test-pipeline.yaml

@@ -314,6 +314,16 @@ steps:
     - export VLLM_WORKER_MULTIPROC_METHOD=spawn
     - pytest -v -s -x lora/test_long_context.py
 
+- label: Weight Loading Multiple GPU Test
+  working_dir: "/vllm-workspace/tests"
+  num_gpus: 2
+  source_file_dependencies:
+  - vllm/
+  - tests/weight_loading
+  commands:
+    - bash weight_loading/run_model_weight_loading_test.sh
+
+
 ##### multi gpus test #####
 ##### A100 test #####
 

tests/weight_loading/models.txt

@@ -0,0 +1,15 @@
+gptq_marlin, robertgshaw2/zephyr-7b-beta-channelwise-gptq, main
+gptq_marlin, TheBloke/Llama-2-7B-GPTQ, main
+gptq_marlin, TheBloke/TinyLlama-1.1B-Chat-v1.0-GPTQ, main
+gptq_marlin, TheBloke/TinyLlama-1.1B-Chat-v1.0-GPTQ, gptq-8bit--1g-actorder_True
+gptq_marlin, TheBloke/TinyLlama-1.1B-Chat-v1.0-GPTQ, gptq-8bit-32g-actorder_True
+gptq_marlin, TechxGenus/gemma-1.1-2b-it-GPTQ, main
+compressed-tensors, nm-testing/tinyllama-oneshot-w8w8-test-static-shape-change, main
+compressed-tensors, nm-testing/tinyllama-oneshot-w8-channel-a8-tensor, main
+compressed-tensors, nm-testing/tinyllama-oneshot-w8a8-dynamic-token-v2, main
+compressed-tensors, nm-testing/tinyllama-oneshot-w8a8-channel-dynamic-token-v2, main
+compressed-tensors, nm-testing/tinyllama-oneshot-w4a16-group128-v2, main
+compressed-tensors, nm-testing/tinyllama-oneshot-w8a16-per-channel, main
+compressed-tensors, nm-testing/Meta-Llama-3-8B-FP8-compressed-tensors-test, main
+compressed-tensors, nm-testing/Phi-3-mini-128k-instruct-FP8, main
+compressed-tensors, neuralmagic/Phi-3-medium-128k-instruct-quantized.w4a16, main

tests/weight_loading/run_model_weight_loading_test.sh

@@ -0,0 +1,32 @@
+#!/bin/bash
+SUCCESS=0
+
+IFS=$'\n' read -d '' -r -a MODEL_CONFIGS < "weight_loading/models.txt"
+
+for MODEL_CONFIG in "${MODEL_CONFIGS[@]}"
+do
+    LOCAL_SUCCESS=0
+    IFS=', ' read -r -a array <<< "$MODEL_CONFIG"
+    
+    echo "=== RUNNING MODEL: $MODEL_CONFIG ==="
+
+    export QUANTIZATION=${array[0]}
+    export MODEL_NAME=${array[1]}
+    export REVISION=${array[2]}
+    pytest -s weight_loading/test_weight_loading.py || LOCAL_SUCCESS=$?
+
+    if [[ $LOCAL_SUCCESS == 0 ]]; then
+        echo "=== PASSED MODEL: ${MODEL_CONFIG} ==="
+    else
+        echo "=== FAILED MODEL: ${MODEL_CONFIG} ==="
+    fi
+
+    SUCCESS=$((SUCCESS + LOCAL_SUCCESS))
+
+done
+
+if [ "${SUCCESS}" -eq "0" ]; then
+    exit 0
+else
+    exit 1
+fi

tests/weight_loading/test_weight_loading.py

@@ -0,0 +1,20 @@
+import os
+
+MAX_MODEL_LEN = 1024
+MODEL_NAME = os.environ.get("MODEL_NAME",
+                            "robertgshaw2/zephyr-7b-beta-channelwise-gptq")
+REVISION = os.environ.get("REVISION", "main")
+QUANTIZATION = os.environ.get("QUANTIZATION", "gptq_marlin")
+
+
+def test_weight_loading(vllm_runner):
+    with vllm_runner(model_name=MODEL_NAME,
+                     revision=REVISION,
+                     dtype="auto",
+                     quantization=QUANTIZATION,
+                     max_model_len=MAX_MODEL_LEN,
+                     tensor_parallel_size=2) as model:
+
+        output = model.generate_greedy("Hello world!", max_tokens=20)
+        print(output)
+        assert output

vllm/model_executor/layers/linear.py

@@ -20,7 +20,9 @@ from vllm.model_executor.utils import set_weight_attrs
 
 logger = init_logger(__name__)
 
-WEIGHT_LOADER_V2_SUPPORTED = ["CompressedTensorsLinearMethod"]
+WEIGHT_LOADER_V2_SUPPORTED = [
+    "CompressedTensorsLinearMethod", "GPTQMarlinLinearMethod"
+]
 
 
 def adjust_marlin_shard(param, shard_size, shard_offset):

vllm/model_executor/layers/quantization/compressed_tensors/schemes/compressed_tensors_wNa16.py

@@ -105,7 +105,7 @@ class CompressedTensorsWNA16(CompressedTensorsScheme):
                 dtype=params_dtype,
             )
         }
-        if self.group_size == -1:
+        if not partition_scales:
             weight_scale = ChannelQuantScaleParameter(output_dim=0,
                                                       **weight_scale_args)
         else:

vllm/model_executor/layers/quantization/gptq_marlin.py

@@ -1,12 +1,11 @@
 from typing import Any, Dict, List, Optional
 
 import torch
-from torch.nn.parameter import Parameter
+from torch.nn import Parameter
 
 from vllm import _custom_ops as ops
 from vllm.logger import init_logger
-from vllm.model_executor.layers.linear import (LinearBase, LinearMethodBase,
-                                               set_weight_attrs)
+from vllm.model_executor.layers.linear import LinearBase, LinearMethodBase
 from vllm.model_executor.layers.quantization.base_config import (
     QuantizationConfig)
 from vllm.model_executor.layers.quantization.utils.marlin_utils import (
@@ -15,6 +14,11 @@ from vllm.model_executor.layers.quantization.utils.marlin_utils import (
     marlin_repeat_scales_on_all_ranks, marlin_sort_g_idx, replace_tensor,
     verify_marlin_supported, verify_marlin_supports_shape)
 from vllm.model_executor.layers.vocab_parallel_embedding import ParallelLMHead
+from vllm.model_executor.parameter import (ChannelQuantScaleParameter,
+                                           GroupQuantScaleParameter,
+                                           PackedColumnParameter,
+                                           PackedvLLMParameter,
+                                           RowvLLMParameter)
 from vllm.scalar_type import scalar_types
 
 logger = init_logger(__name__)
@@ -159,9 +163,11 @@ class GPTQMarlinLinearMethod(LinearMethodBase):
         params_dtype: torch.dtype,
         **extra_weight_attrs,
     ) -> None:
+
         del output_size
         output_size_per_partition = sum(output_partition_sizes)
         is_row_parallel = input_size != input_size_per_partition
+        weight_loader = extra_weight_attrs.get("weight_loader")
 
         # Normalize group_size
         if self.quant_config.group_size != -1:
@@ -190,79 +196,66 @@ class GPTQMarlinLinearMethod(LinearMethodBase):
             scales_and_zp_size = input_size_per_partition // group_size
 
         # Quantized weights
-        qweight = Parameter(
-            torch.empty(
+        qweight = PackedvLLMParameter(
+            data=torch.empty(
                 input_size_per_partition // self.quant_config.pack_factor,
                 output_size_per_partition,
                 dtype=torch.int32,
             ),
-            requires_grad=False,
-        )
-        set_weight_attrs(
-            qweight,
-            {
-                **extra_weight_attrs,
-                "input_dim": 0,
-                "output_dim": 1,
-                "packed_dim": 0,
-                "pack_factor": self.quant_config.pack_factor,
-            },
-        )
+            input_dim=0,
+            output_dim=1,
+            packed_dim=0,
+            packed_factor=self.quant_config.pack_factor,
+            weight_loader=weight_loader)
 
         # Activation order
-        g_idx = Parameter(
-            torch.empty(
-                input_size_per_partition,
-                dtype=torch.int32,
-            ),
-            requires_grad=False,
-        )
-        # Ignore warning from fused linear layers such as QKVParallelLinear.
-        set_weight_attrs(
-            g_idx,
-            {
-                **extra_weight_attrs, "input_dim": 0,
-                "ignore_warning": True
-            },
-        )
+        g_idx = RowvLLMParameter(data=torch.empty(
+            input_size_per_partition,
+            dtype=torch.int32,
+        ),
+                                 input_dim=0,
+                                 weight_loader=weight_loader)
 
-        # Scales
-        scales = Parameter(
-            torch.empty(
-                scales_and_zp_size,
-                output_size_per_partition,
-                dtype=params_dtype,
-            ),
-            requires_grad=False,
-        )
-        set_weight_attrs(
-            scales,
-            {
-                **extra_weight_attrs,
-                "input_dim": scales_and_zp_input_dim,
-                "output_dim": 1,
-            },
-        )
-
-        # Quantized zero-points
-        qzeros = Parameter(
+        qzeros_args = {
+            "data":
             torch.empty(
                 scales_and_zp_size,
                 output_size_per_partition // self.quant_config.pack_factor,
                 dtype=torch.int32,
             ),
-            requires_grad=False,
-        )
-        set_weight_attrs(
-            qzeros,
-            {
-                **extra_weight_attrs,
-                "input_dim": scales_and_zp_input_dim,
-                "output_dim": 1,
-                "packed_dim": 1,
-                "pack_factor": self.quant_config.pack_factor,
-            },
-        )
+            "weight_loader":
+            weight_loader
+        }
+        weight_scale_args = {
+            "data":
+            torch.empty(
+                scales_and_zp_size,
+                output_size_per_partition,
+                dtype=params_dtype,
+            ),
+            "weight_loader":
+            weight_loader
+        }
+
+        if scales_and_zp_input_dim is None:
+            scales = ChannelQuantScaleParameter(output_dim=1,
+                                                **weight_scale_args)
+            qzeros = PackedColumnParameter(
+                output_dim=1,
+                packed_dim=1,
+                packed_factor=self.quant_config.pack_factor,
+                **qzeros_args)
+
+        else:
+            scales = GroupQuantScaleParameter(output_dim=1,
+                                              input_dim=0,
+                                              **weight_scale_args)
+            qzeros = PackedvLLMParameter(
+                input_dim=0,
+                output_dim=1,
+                packed_dim=1,
+                packed_factor=self.quant_config.pack_factor,
+                **qzeros_args)
 
         layer.register_parameter("qweight", qweight)
         layer.register_parameter("g_idx", g_idx)
@@ -280,6 +273,10 @@ class GPTQMarlinLinearMethod(LinearMethodBase):
     def process_weights_after_loading(self, layer: torch.nn.Module) -> None:
         device = layer.qweight.device
 
+        # required by torch.compile
+        layer.qweight = Parameter(layer.qweight.data, requires_grad=False)
+        layer.scales = Parameter(layer.scales.data, requires_grad=False)
+
         # Allocate marlin workspace
         layer.workspace = marlin_make_workspace(
             layer.output_size_per_partition, device)

vllm/model_executor/parameter.py

@@ -9,7 +9,7 @@ from vllm.logger import init_logger
 __all__ = [
     "BasevLLMParameter", "PackedvLLMParameter", "PerTensorScaleParameter",
     "ModelWeightParameter", "ChannelQuantScaleParameter",
-    "GroupQuantScaleParameter"
+    "GroupQuantScaleParameter", "PackedColumnParameter", "RowvLLMParameter"
 ]
 
 logger = init_logger(__name__)
@@ -92,7 +92,8 @@ class _ColumnvLLMParameter(BasevLLMParameter):
         shard_size = kwargs.get("shard_size")
         if isinstance(
                 self,
-                PackedvLLMParameter) and self.packed_dim == self.output_dim:
+            (PackedColumnParameter,
+             PackedvLLMParameter)) and self.packed_dim == self.output_dim:
             shard_size, shard_offset = self.adjust_shard_indexes_for_packing(
                 shard_offset=shard_offset, shard_size=shard_size)
 
@@ -115,7 +116,8 @@ class _ColumnvLLMParameter(BasevLLMParameter):
 
         if isinstance(
                 self,
-                PackedvLLMParameter) and self.output_dim == self.packed_dim:
+            (PackedColumnParameter,
+             PackedvLLMParameter)) and self.output_dim == self.packed_dim:
             shard_size, shard_offset = self.adjust_shard_indexes_for_packing(
                 shard_offset=shard_offset, shard_size=shard_size)
 
@@ -131,12 +133,12 @@ class _ColumnvLLMParameter(BasevLLMParameter):
         param_data.copy_(loaded_weight)
 
 
-class ModelWeightParameter(_ColumnvLLMParameter):
+class RowvLLMParameter(BasevLLMParameter):
     """
-    Parameter class for linear layer weights. Extends the
-    _ColumnvLLMParameter by adding loading functionality
-    for linear layers with row parallel functionality.
-    Requires an input dimension to be defined.
+    Parameter class defining weight_loading functionality
+    (load_row_parallel_weight) for parameters being loaded
+    into linear layers with row parallel functionality.
+    Requires an input_dim to be defined.
     """
 
     def __init__(self, input_dim: int, **kwargs):
@@ -160,10 +162,18 @@ class ModelWeightParameter(_ColumnvLLMParameter):
         self.data.copy_(loaded_weight)
 
 
-class GroupQuantScaleParameter(ModelWeightParameter):
+class ModelWeightParameter(_ColumnvLLMParameter, RowvLLMParameter):
+    """
+    Parameter class for linear layer weights. Uses both column and
+    row parallelism.
+    """
+    pass
+
+
+class GroupQuantScaleParameter(_ColumnvLLMParameter, RowvLLMParameter):
     """
     Parameter class for weight scales loaded for weights with
-    grouped quantization. Equivalent to ModelWeightParameter.
+    grouped quantization. Uses both column and row parallelism.
     """
     pass
 
@@ -171,7 +181,7 @@ class GroupQuantScaleParameter(ModelWeightParameter):
 class ChannelQuantScaleParameter(_ColumnvLLMParameter):
     """
     Parameter class for weight scales loaded for weights with
-    channel-wise quantization. Equivalent to _ColumnvLLMParameter. 
+    channel-wise quantization. Equivalent to _ColumnvLLMParameter.
     """
     pass
 
@@ -181,7 +191,7 @@ class PerTensorScaleParameter(BasevLLMParameter):
     Parameter class for scales where the number of scales is
     equivalent to the number of logical matrices in fused linear
     layers (e.g. for QKV, there are 3 scales loaded from disk).
-    This is relevant to weights with per-tensor quantization. 
+    This is relevant to weights with per-tensor quantization.
     Adds functionality to map the scalers to a shard during
     weight loading. 
 
@@ -232,15 +242,11 @@ class PerTensorScaleParameter(BasevLLMParameter):
         param_data.copy_(loaded_weight)
 
 
-class PackedvLLMParameter(ModelWeightParameter):
+class PackedColumnParameter(_ColumnvLLMParameter):
     """
-    Parameter for model weights which are packed on disk.
-    Example: GPTQ Marlin weights are int4 or int8, packed into int32.
-    Extends the ModelWeightParameter to take in the
-    packed factor, the packed dimension, and optionally, marlin
-    tile size for marlin kernels. Adjusts the shard_size and 
-    shard_offset for fused linear layers model weight loading 
-    by accounting for packing and optionally, marlin tile size.
+    Parameter for model parameters which are packed on disk
+    and support column parallelism only. See PackedvLLMParameter
+    for more details on the packed properties.
     """
 
     def __init__(self,
@@ -250,7 +256,7 @@ class PackedvLLMParameter(ModelWeightParameter):
                  **kwargs):
         self._packed_factor = packed_factor
         self._packed_dim = packed_dim
-        self._marlin_tile = marlin_tile_size
+        self._marlin_tile_size = marlin_tile_size
         super().__init__(**kwargs)
 
     @property
@@ -262,16 +268,70 @@ class PackedvLLMParameter(ModelWeightParameter):
         return self._packed_factor
 
     @property
-    def marlin_tile(self):
-        return self._marlin_tile
-
-    def _adjust_shard_indexes_for_marlin(self, shard_size, shard_offset):
-        return shard_size * self.marlin_tile, shard_offset * self.marlin_tile
+    def marlin_tile_size(self):
+        return self._marlin_tile_size
 
     def adjust_shard_indexes_for_packing(self, shard_size, shard_offset):
-        shard_size = shard_size // self.packed_factor
-        shard_offset = shard_offset // self.packed_factor
-        if self.marlin_tile is not None:
-            return self._adjust_shard_indexes_for_marlin(
-                shard_size, shard_offset)
-        return shard_size, shard_offset
+        return _adjust_shard_indexes_for_packing(
+            shard_size=shard_size,
+            shard_offset=shard_offset,
+            packed_factor=self.packed_factor,
+            marlin_tile_size=self.marlin_tile_size)
+
+
+class PackedvLLMParameter(ModelWeightParameter):
+    """
+    Parameter for model weights which are packed on disk.
+    Example: GPTQ Marlin weights are int4 or int8, packed into int32.
+    Extends the ModelWeightParameter to take in the
+    packed factor, the packed dimension, and optionally, marlin
+    tile size for marlin kernels. Adjusts the shard_size and 
+    shard_offset for fused linear layers model weight loading
+    by accounting for packing and optionally, marlin tile size.
+    """
+
+    def __init__(self,
+                 packed_factor: int,
+                 packed_dim: int,
+                 marlin_tile_size: Optional[int] = None,
+                 **kwargs):
+        self._packed_factor = packed_factor
+        self._packed_dim = packed_dim
+        self._marlin_tile_size = marlin_tile_size
+        super().__init__(**kwargs)
+
+    @property
+    def packed_dim(self):
+        return self._packed_dim
+
+    @property
+    def packed_factor(self):
+        return self._packed_factor
+
+    @property
+    def marlin_tile_size(self):
+        return self._marlin_tile_size
+
+    def adjust_shard_indexes_for_packing(self, shard_size, shard_offset):
+        return _adjust_shard_indexes_for_packing(
+            shard_size=shard_size,
+            shard_offset=shard_offset,
+            packed_factor=self.packed_factor,
+            marlin_tile_size=self.marlin_tile_size)
+
+
+def _adjust_shard_indexes_for_marlin(shard_size, shard_offset,
+                                     marlin_tile_size):
+    return shard_size * marlin_tile_size, shard_offset * marlin_tile_size
+
+
+def _adjust_shard_indexes_for_packing(shard_size, shard_offset, packed_factor,
+                                      marlin_tile_size):
+    shard_size = shard_size // packed_factor
+    shard_offset = shard_offset // packed_factor
+    if marlin_tile_size is not None:
+        return _adjust_shard_indexes_for_marlin(
+            shard_size=shard_size,
+            shard_offset=shard_offset,
+            marlin_tile_size=marlin_tile_size)
+    return shard_size, shard_offset