Fix quantization and nightly tests (#4258)

python/sglang/srt/layers/moe/fused_moe_triton/fused_moe.py

@@ -23,7 +23,6 @@ from sglang.srt.utils import (
     direct_register_custom_op,
     get_bool_env_var,
     get_device_name,
-    is_cuda_available,
     is_hip,
 )
 

python/sglang/srt/layers/quantization/__init__.py

 # Adapted from https://raw.githubusercontent.com/vllm-project/vllm/v0.5.5/vllm/model_executor/layers/quantization/__init__.py
+import builtins
+import inspect
 import re
 from copy import deepcopy
 from typing import Callable, Dict, Optional, Type, Union
@@ -6,10 +8,7 @@ from typing import Callable, Dict, Optional, Type, Union
 import torch
 from vllm.model_executor.layers.quantization.aqlm import AQLMConfig
 from vllm.model_executor.layers.quantization.awq import AWQConfig
-from vllm.model_executor.layers.quantization.awq_marlin import (
-    AWQMarlinConfig,
-    AWQMoEMethod,
-)
+from vllm.model_executor.layers.quantization.awq_marlin import AWQMarlinConfig
 from vllm.model_executor.layers.quantization.bitsandbytes import BitsAndBytesConfig
 from vllm.model_executor.layers.quantization.compressed_tensors.compressed_tensors import (
     CompressedTensorsConfig,
@@ -180,96 +179,117 @@ def gptq_get_quant_method(self, layer, prefix):
     return None
 
 
-def awq_get_quant_method(self, layer, prefix):
-    from vllm.model_executor.layers.quantization.awq import is_layer_skipped_awq
-    from vllm.model_executor.layers.quantization.awq_marlin import (
-        AWQMarlinLinearMethod,
-        AWQMoEMethod,
+original_isinstance = builtins.isinstance
+
+
+def monkey_patch_isinstance_for_vllm_base_layer(reverse: bool = False):
+    """
+    Patch isinstance so that the `get_quant_method` in vllm's QuantizationConfig
+    can recognize sglang layers
+    """
+
+    if reverse:
+        builtins.isinstance = original_isinstance
+        return
+
+    from vllm.model_executor.layers.fused_moe import FusedMoE
+    from vllm.model_executor.layers.linear import LinearBase
+    from vllm.model_executor.layers.vocab_parallel_embedding import (
+        VocabParallelEmbedding,
     )
 
-    from sglang.srt.layers.linear import LinearBase, UnquantizedLinearMethod
-    from sglang.srt.layers.moe.fused_moe_triton.layer import FusedMoE
-    from sglang.srt.layers.vocab_parallel_embedding import ParallelLMHead
+    from sglang.srt.layers.linear import LinearBase as PatchedLinearBase
+    from sglang.srt.layers.moe.fused_moe_triton.layer import FusedMoE as PatchedFusedMoE
+    from sglang.srt.layers.vocab_parallel_embedding import (
+        VocabParallelEmbedding as PatchedVocabParallelEmbedding,
+    )
 
-    if isinstance(layer, LinearBase) or (
-        isinstance(layer, ParallelLMHead) and self.lm_head_quantized
-    ):
-        if is_layer_skipped_awq(prefix, self.modules_to_not_convert):
-            return UnquantizedLinearMethod()
-        return AWQMarlinLinearMethod(self)
-    elif isinstance(layer, FusedMoE):
-        return AWQMoEMethod(self)
-    return None
+    def patched_isinstance(obj, classinfo):
+        if classinfo is LinearBase:
+            return original_isinstance(obj, PatchedLinearBase)
+        if classinfo is FusedMoE:
+            return original_isinstance(obj, PatchedFusedMoE)
+        if classinfo is VocabParallelEmbedding:
+            return original_isinstance(obj, PatchedVocabParallelEmbedding)
+        return original_isinstance(obj, classinfo)
 
+    builtins.isinstance = patched_isinstance
 
-original_awq_moe_method_apply = AWQMoEMethod.apply
 
+def monkey_patch_moe_apply(class_obj: "FusedMoEMethodBase"):
+    """
+    Monkey patch the apply function of vllm's FusedMoEMethodBase.
+    Convert sglang arguments to vllm arguments.
+    """
+    original_apply = class_obj.apply
+    sig = inspect.signature(original_apply)
+    param_names = list(sig.parameters.keys())
+    has_correction_bias = "e_score_correction_bias" in param_names
 
-def awq_moe_method_apply(
+    def new_apply(
         self,
         layer: torch.nn.Module,
         x: torch.Tensor,
         router_logits: torch.Tensor,
         top_k: int,
         renormalize: bool,
-    use_grouped_topk: bool = False,
+        use_grouped_topk: bool,
         topk_group: Optional[int] = None,
         num_expert_group: Optional[int] = None,
         custom_routing_function: Optional[Callable] = None,
-    scoring_func: str = "softmax",
-    e_score_correction_bias: Optional[torch.Tensor] = None,
-    **kwargs,
-):
-    return original_awq_moe_method_apply(
-        self,
-        layer,
-        x,
-        router_logits,
-        top_k,
-        renormalize,
-        use_grouped_topk,
-        topk_group,
-        num_expert_group,
-        custom_routing_function,
-        scoring_func,
-        e_score_correction_bias,
+        correction_bias: Optional[torch.Tensor] = None,
+        activation: str = "silu",
+        inplace: bool = True,
+        no_combine: bool = False,
+    ):
+        assert activation == "silu"
+        assert inplace and not no_combine
+
+        kwargs = {
+            "self": self,
+            "layer": layer,
+            "x": x,
+            "router_logits": router_logits,
+            "top_k": top_k,
+            "renormalize": renormalize,
+            "use_grouped_topk": use_grouped_topk,
+            "topk_group": topk_group,
+            "num_expert_group": num_expert_group,
+            "custom_routing_function": custom_routing_function,
+        }
+        if correction_bias is not None:
+            if not has_correction_bias:
+                raise ValueError(
+                    "Please increase the version of your vllm. Try `pip install vllm==0.7.2`"
                 )
+            kwargs["e_score_correction_bias"] = correction_bias
+        return original_apply(**kwargs)
 
+    setattr(class_obj, "apply", new_apply)
 
-def patch_vllm_linear_base_isinstance():
-    import builtins
-
-    from vllm.model_executor.layers.linear import LinearBase
-
-    from sglang.srt.layers.linear import LinearBase as PatchedLinearBase
 
-    original_isinstance = builtins.isinstance
-
-    def patched_isinstance(obj, classinfo):
-        if classinfo is LinearBase:
-            return original_isinstance(obj, PatchedLinearBase)
-        return original_isinstance(obj, classinfo)
-
-    builtins.isinstance = patched_isinstance
-
-
-def apply_monkey_patches():
+def monkey_patch_quant_configs():
     """Apply all monkey patches in one place."""
     from vllm.model_executor.layers.quantization.awq_marlin import AWQMoEMethod
+    from vllm.model_executor.layers.quantization.compressed_tensors.compressed_tensors_moe import (
+        CompressedTensorsW8A8Fp8MoEMethod,
+        CompressedTensorsWNA16MoEMethod,
+    )
+    from vllm.model_executor.layers.quantization.gptq_marlin import GPTQMarlinMoEMethod
 
     setattr(GPTQMarlinConfig, "get_quant_method", gptq_get_quant_method)
     setattr(GPTQConfig, "get_quant_method", gptq_get_quant_method)
-    setattr(AWQMarlinConfig, "get_quant_method", awq_get_quant_method)
-    setattr(AWQMoEMethod, "apply", awq_moe_method_apply)
+
+    monkey_patch_moe_apply(AWQMoEMethod)
+    monkey_patch_moe_apply(GPTQMarlinMoEMethod)
+    monkey_patch_moe_apply(CompressedTensorsW8A8Fp8MoEMethod)
+    monkey_patch_moe_apply(CompressedTensorsWNA16MoEMethod)
 
 
-patch_vllm_linear_base_isinstance()
-# Apply patches when module is imported
-apply_monkey_patches()
+monkey_patch_quant_configs()
 
 
 __all__ = [
-    "QuantizationConfig",
     "get_quantization_config",
     "QUANTIZATION_METHODS",
 ]

python/sglang/srt/model_executor/model_runner.py

@@ -41,6 +41,7 @@ from sglang.srt.layers.dp_attention import (
     initialize_dp_attention,
 )
 from sglang.srt.layers.logits_processor import LogitsProcessorOutput
+from sglang.srt.layers.quantization import monkey_patch_isinstance_for_vllm_base_layer
 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
@@ -341,6 +342,8 @@ class ModelRunner:
         # Load the model
         # Remove monkey_patch when linear.py quant remove dependencies with vllm
         monkey_patch_vllm_parallel_state()
+        monkey_patch_isinstance_for_vllm_base_layer()
+
         with self.memory_saver_adapter.region():
             self.model = get_model(
                 model_config=self.model_config,
@@ -348,6 +351,7 @@ class ModelRunner:
                 device_config=DeviceConfig(self.device),
             )
         monkey_patch_vllm_parallel_state(reverse=True)
+        monkey_patch_isinstance_for_vllm_base_layer(reverse=True)
 
         if self.server_args.kv_cache_dtype == "fp8_e4m3":
             if self.server_args.quantization_param_path is not None:

python/sglang/test/test_utils.py

@@ -36,12 +36,15 @@ DEFAULT_SMALL_EMBEDDING_MODEL_NAME_FOR_TEST = "Alibaba-NLP/gte-Qwen2-1.5B-instru
 DEFAULT_MLA_MODEL_NAME_FOR_TEST = "deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct"
 DEFAULT_MLA_FP8_MODEL_NAME_FOR_TEST = "neuralmagic/DeepSeek-Coder-V2-Lite-Instruct-FP8"
 DEFAULT_REASONING_MODEL_NAME_FOR_TEST = "deepseek-ai/DeepSeek-R1-Distill-Qwen-7B"
+DEFAULT_AWQ_MOE_MODEL_NAME_FOR_TEST = (
+    "hugging-quants/Mixtral-8x7B-Instruct-v0.1-AWQ-INT4"
+)
 DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH = 1000
 DEFAULT_MODEL_NAME_FOR_NIGHTLY_EVAL_TP1 = "meta-llama/Llama-3.1-8B-Instruct,mistralai/Mistral-7B-Instruct-v0.3,deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct,google/gemma-2-27b-it"
 DEFAULT_MODEL_NAME_FOR_NIGHTLY_EVAL_TP2 = "meta-llama/Llama-3.1-70B-Instruct,mistralai/Mixtral-8x7B-Instruct-v0.1,Qwen/Qwen2-57B-A14B-Instruct"
 DEFAULT_MODEL_NAME_FOR_NIGHTLY_EVAL_FP8_TP1 = "neuralmagic/Meta-Llama-3.1-8B-Instruct-FP8,neuralmagic/Mistral-7B-Instruct-v0.3-FP8,neuralmagic/DeepSeek-Coder-V2-Lite-Instruct-FP8,neuralmagic/gemma-2-2b-it-FP8"
 DEFAULT_MODEL_NAME_FOR_NIGHTLY_EVAL_FP8_TP2 = "neuralmagic/Meta-Llama-3.1-70B-Instruct-FP8,neuralmagic/Mixtral-8x7B-Instruct-v0.1-FP8,neuralmagic/Qwen2-72B-Instruct-FP8,neuralmagic/Qwen2-57B-A14B-Instruct-FP8,neuralmagic/DeepSeek-Coder-V2-Lite-Instruct-FP8"
-DEFAULT_MODEL_NAME_FOR_NIGHTLY_EVAL_QUANT_TP1 = "hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4,hugging-quants/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4"
+DEFAULT_MODEL_NAME_FOR_NIGHTLY_EVAL_QUANT_TP1 = "hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4,hugging-quants/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4,hugging-quants/Mixtral-8x7B-Instruct-v0.1-AWQ-INT4"
 DEFAULT_SMALL_MODEL_NAME_FOR_TEST_QWEN = "Qwen/Qwen2.5-1.5B-Instruct"
 DEFAULT_SMALL_VLM_MODEL_NAME = "Qwen/Qwen2-VL-2B"
 

test/srt/run_suite.py

@@ -22,6 +22,7 @@ suites = {
         TestFile("models/test_reward_models.py", 83),
         TestFile("models/test_gme_qwen_models.py", 45),
         TestFile("test_abort.py", 51),
+        TestFile("test_awq.py"),
         TestFile("test_block_int8.py", 22),
         TestFile("test_chunked_prefill.py", 336),
         TestFile("test_eagle_infer.py", 447),

test/srt/test_awq.py

+import unittest
+from types import SimpleNamespace
+
+from sglang.srt.utils import kill_process_tree
+from sglang.test.run_eval import run_eval
+from sglang.test.test_utils import (
+    DEFAULT_AWQ_MOE_MODEL_NAME_FOR_TEST,
+    DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+    DEFAULT_URL_FOR_TEST,
+    popen_launch_server,
+)
+
+
+class TestAWQ(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.model = DEFAULT_AWQ_MOE_MODEL_NAME_FOR_TEST
+        cls.base_url = DEFAULT_URL_FOR_TEST
+        cls.process = popen_launch_server(
+            cls.model,
+            cls.base_url,
+            timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+            other_args=["--trust-remote-code"],
+        )
+
+    @classmethod
+    def tearDownClass(cls):
+        kill_process_tree(cls.process.pid)
+
+    def test_mmlu(self):
+        args = SimpleNamespace(
+            base_url=self.base_url,
+            model=self.model,
+            eval_name="mmlu",
+            num_examples=64,
+            num_threads=32,
+        )
+
+        metrics = run_eval(args)
+        self.assertGreater(metrics["score"], 0.65)
+
+
+if __name__ == "__main__":
+    unittest.main()

test/srt/test_nightly_gsm8k_eval.py

@@ -38,6 +38,7 @@ MODEL_SCORE_THRESHOLDS = {
     "neuralmagic/Qwen2-57B-A14B-Instruct-FP8": 0.82,
     "hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4": 0.84,
     "hugging-quants/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4": 0.83,
+    "hugging-quants/Mixtral-8x7B-Instruct-v0.1-AWQ-INT4": 0.60,
 }