[Hardware][Ascend] Add Ascend NPU backend

Dockerfile.npu

@@ -0,0 +1,19 @@
+FROM ascendai/cann:8.0.rc3.alpha002-910b-ubuntu22.04-py3.9
+## Change to use the following if you failed with the above one
+# FROM quay.io/ascend/cann:8.0.rc3.alpha002-910b-ubuntu22.04-py3.9
+
+# Define environments
+ENV DEBIAN_FRONTEND=noninteractive
+
+RUN apt-get update -y && \
+    apt-get install -y python3-pip git vim
+WORKDIR /workspace
+
+COPY . /workspace/vllm/
+
+# install build requirements
+RUN PIP_EXTRA_INDEX_URL="https://download.pytorch.org/whl/cpu" python3 -m pip install -r /workspace/vllm/requirements-build.txt
+# build vLLM with NPU backend
+RUN PIP_EXTRA_INDEX_URL="https://download.pytorch.org/whl/cpu" VLLM_TARGET_DEVICE="npu" python3 -m pip install /workspace/vllm/
+
+CMD ["/bin/bash"]

examples/offline_inference_npu.py

@@ -0,0 +1,41 @@
+import gc
+
+import torch
+
+from vllm import LLM, SamplingParams
+from vllm.distributed.parallel_state import (destroy_distributed_environment,
+                                             destroy_model_parallel)
+
+
+def clean_up():
+    destroy_model_parallel()
+    destroy_distributed_environment()
+    gc.collect()
+    torch.npu.empty_cache()
+
+
+# Sample prompts.
+prompts = [
+    "Hello, my name is",
+    "The president of the United States is",
+    "The capital of France is",
+    "The future of AI is",
+]
+
+# Create a sampling params object.
+sampling_params = SamplingParams(max_tokens=100, temperature=0.0)
+
+# Create an LLM.
+llm = LLM(model="facebook/opt-125m")
+
+# Generate texts from the prompts. The output is a list of RequestOutput objects
+# that contain the prompt, generated text, and other information.
+outputs = llm.generate(prompts, sampling_params)
+# Print the outputs.
+for output in outputs:
+    prompt = output.prompt
+    generated_text = output.outputs[0].text
+    print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")
+
+del llm
+clean_up()

pyproject.toml

@@ -83,7 +83,7 @@ exclude = [
 ]
 
 [tool.codespell]
-ignore-words-list = "dout, te, indicies, subtile"
+ignore-words-list = "dout, te, indicies, subtile, cann"
 skip = "./tests/models/fixtures,./tests/prompts,./benchmarks/sonnet.txt,./tests/lora/data,./build"
 
 [tool.isort]

requirements-npu.txt

@@ -0,0 +1,8 @@
+# Common dependencies
+-r requirements-common.txt
+
+decorator
+pyyaml
+scipy
+setuptools
+torch_npu == 2.5.1rc1

setup.py

@@ -307,6 +307,10 @@ def _is_xpu() -> bool:
     return VLLM_TARGET_DEVICE == "xpu"
 
 
+def _is_npu() -> bool:
+    return VLLM_TARGET_DEVICE == "npu"
+
+
 def _build_custom_ops() -> bool:
     return _is_cuda() or _is_hip() or _is_cpu()
 
@@ -429,6 +433,8 @@ def get_vllm_version() -> str:
         version += f"{sep}cpu"
     elif _is_xpu():
         version += f"{sep}xpu"
+    elif _is_npu():
+        version += f"{sep}npu"
     else:
         raise RuntimeError("Unknown runtime environment")
 
@@ -489,6 +495,8 @@ def _read_requirements(filename: str) -> List[str]:
         requirements = _read_requirements("requirements-cpu.txt")
     elif _is_xpu():
         requirements = _read_requirements("requirements-xpu.txt")
+    elif _is_npu():
+        requirements = _read_requirements("requirements-npu.txt")
     else:
         raise ValueError(
             "Unsupported platform, please use CUDA, ROCm, Neuron, HPU, "

vllm/_custom_ops.py

@@ -13,7 +13,8 @@
 
 logger = init_logger(__name__)
 
-if not current_platform.is_tpu() and not current_platform.is_hpu():
+if not (current_platform.is_tpu() or current_platform.is_hpu()
+        or current_platform.is_npu()):
     try:
         import vllm._C
     except ImportError as e: