model_loader.py

"""Utilities for selecting and loading models."""
import contextlib
from typing import Type

import torch
import torch.nn as nn

from vllm.config import DeviceConfig, ModelConfig
from vllm.model_executor.models import ModelRegistry
from vllm.model_executor.weight_utils import (get_quant_config,
                                              initialize_dummy_weights)
import os


@contextlib.contextmanager
def _set_default_torch_dtype(dtype: torch.dtype):
    """Sets the default torch dtype to the given dtype."""
    old_dtype = torch.get_default_dtype()
    torch.set_default_dtype(dtype)
    yield
    torch.set_default_dtype(old_dtype)


def _get_model_architecture(model_config: ModelConfig) -> Type[nn.Module]:
    architectures = getattr(model_config.hf_config, "architectures", [])
    if architectures == ['LlamaForCausalLM'] or architectures == ['ChatGLMModel'] or architectures == ['BaichuanForCausalLM']:
        if os.getenv('LLAMA_NN') != '0': 
            os.environ['LLAMA_NN'] = '1'
    # Special handling for quantized Mixtral.
    # FIXME(woosuk): This is a temporary hack.
    if (model_config.quantization is not None
            and "MixtralForCausalLM" in architectures):
        architectures = ["QuantMixtralForCausalLM"]

    for arch in architectures:
        model_cls = ModelRegistry.load_model_cls(arch)
        if model_cls is not None:
            return model_cls
    raise ValueError(
        f"Model architectures {architectures} are not supported for now. "
        f"Supported architectures: {ModelRegistry.get_supported_archs()}")


def get_model(model_config: ModelConfig, device_config: DeviceConfig,
              **kwargs) -> nn.Module:
    lora_config = kwargs.get("lora_config", None)
    model_class = _get_model_architecture(model_config)

    # Get the (maybe quantized) linear method.
    linear_method = None
    if model_config.quantization is not None:
        quant_config = get_quant_config(model_config)
        capability = torch.cuda.get_device_capability()
        capability = capability[0] * 10 + capability[1]
        if capability < quant_config.get_min_capability():
            raise ValueError(
                f"The quantization method {model_config.quantization} is not "
                "supported for the current GPU. "
                f"Minimum capability: {quant_config.get_min_capability()}. "
                f"Current capability: {capability}.")
        supported_dtypes = quant_config.get_supported_act_dtypes()
        if model_config.dtype not in supported_dtypes:
            raise ValueError(
                f"{model_config.dtype} is not supported for quantization "
                f"method {model_config.quantization}. Supported dtypes: "
                f"{supported_dtypes}")
        linear_method = quant_config.get_linear_method()

    if linear_method != None:
        os.environ['LLAMA_NN'] = '0'
        
    with _set_default_torch_dtype(model_config.dtype):
        # Create a model instance.
        # The weights will be initialized as empty tensors.
        with torch.device(device_config.device):
            if hasattr(model_class, "supported_lora_modules"):
                model = model_class(model_config.hf_config, linear_method,
                                    lora_config)
            elif lora_config:
                raise ValueError(
                    f"Model {model_class.__name__} does not support LoRA, "
                    "but LoRA is enabled. Support for this model may "
                    "be added in the future. If this is important to you, "
                    "please open an issue on github.")
            else:
                model = model_class(model_config.hf_config, linear_method)
        if model_config.load_format == "dummy":
            # NOTE(woosuk): For accurate performance evaluation, we assign
            # random values to the weights.
            initialize_dummy_weights(model)
        else:
            # Load the weights from the cached or downloaded files.
            model.load_weights(model_config.model, model_config.download_dir,
                               model_config.load_format, model_config.revision)
    return model.eval()