src/transformers/integrations/awq.py

# Copyright 2023 The HuggingFace Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"AWQ (Activation aware Weight Quantization) integration file"
from ..activations import ACT2FN
from ..modeling_utils import PreTrainedModel
from ..utils import is_auto_awq_available, is_torch_available
from ..utils.quantization_config import (
    AwqBackendPackingMethod,
    AwqConfig,
    AWQLinearVersion,
    ExllamaVersion,
)


if is_torch_available():
    import torch
    import torch.nn as nn


AWQ_FUSED_MAPPINGS = {
    "mistral": {
        "attention": ["q_proj", "k_proj", "v_proj", "o_proj"],
        "mlp": ["gate_proj", "up_proj", "down_proj"],
        "layernorm": ["input_layernorm", "post_attention_layernorm", "norm"],
        "use_alibi": False,
    },
    "mixtral": {
        "attention": ["q_proj", "k_proj", "v_proj", "o_proj"],
        "mlp": ["w1", "w3", "w2"],
        "layernorm": ["input_layernorm", "post_attention_layernorm", "norm"],
        "use_alibi": False,
        "rope_theta": 1000000.0,
    },
    "llama": {
        "attention": ["q_proj", "k_proj", "v_proj", "o_proj"],
        "mlp": ["gate_proj", "up_proj", "down_proj"],
        "layernorm": ["input_layernorm", "post_attention_layernorm", "norm"],
        "use_alibi": False,
    },
    "llava": {
        "attention": ["q_proj", "k_proj", "v_proj", "o_proj"],
        "mlp": ["gate_proj", "up_proj", "down_proj"],
        "layernorm": ["input_layernorm", "post_attention_layernorm", "norm"],
        "use_alibi": False,
    },
}


def replace_with_awq_linear(
    model,
    modules_to_not_convert=None,
    quantization_config=None,
    current_key_name=None,
    has_been_replaced=False,
) -> bool:
    """
    Public method that recursively replaces the Linear layers of the given model with AWQ quantized layers.
    `accelerate` is needed to use this method. Returns the converted model and a boolean that indicates if the
    conversion has been successfull or not.

    During the module replacement, we also infer the backend to use through the `quantization_config` object.

    Args:
        model (`torch.nn.Module`):
            The model to convert, can be any `torch.nn.Module` instance.
        quantization_config (`AwqConfig`):
            The quantization config object that contains the quantization parameters.
        modules_to_not_convert (`list`, *optional*):
            A list of modules to not convert. If a module name is in the list (e.g. `lm_head`), it will not be
            converted.
        current_key_name (`list`, *optional*):
            A list that contains the current key name. This is used for recursion and should not be passed by the user.
        has_been_replaced (`bool`, *optional*):
            A boolean that indicates if the conversion has been successful or not. This is used for recursion and
            should not be passed by the user.
    """
    if modules_to_not_convert is None:
        modules_to_not_convert = []

    backend = quantization_config.backend

    if not is_auto_awq_available():
        raise ValueError(
            "AWQ (either `autoawq` or `llmawq`) is not available. Please install it with `pip install autoawq` or check out the installation guide in https://github.com/mit-han-lab/llm-awq"
        )

    if backend == AwqBackendPackingMethod.AUTOAWQ:
        if quantization_config.version == AWQLinearVersion.GEMM:
            from awq.modules.linear.gemm import WQLinear_GEMM

            target_cls = WQLinear_GEMM
        elif quantization_config.version == AWQLinearVersion.GEMV:
            from awq.modules.linear.gemv import WQLinear_GEMV

            target_cls = WQLinear_GEMV
        elif quantization_config.version == AWQLinearVersion.EXLLAMA:
            if quantization_config.exllama_config["version"] == ExllamaVersion.ONE:
                from awq.modules.linear.exllama import WQLinear_Exllama

                target_cls = WQLinear_Exllama
            elif quantization_config.exllama_config["version"] == ExllamaVersion.TWO:
                from awq.modules.linear.exllamav2 import WQLinear_ExllamaV2

                target_cls = WQLinear_ExllamaV2
            else:
                raise ValueError(f"Unrecognized Exllama version: {quantization_config.exllama_config['version']}")
        else:
            raise ValueError(f"Unrecognized AWQ version: {quantization_config.version}")
    else:
        from awq.quantize.qmodule import WQLinear

        target_cls = WQLinear

    for name, module in model.named_children():
        if current_key_name is None:
            current_key_name = []
        current_key_name.append(name)

        if isinstance(module, nn.Linear) and name not in modules_to_not_convert:
            # Check if the current key is not in the `modules_to_not_convert`
            if not any(key in ".".join(current_key_name) for key in modules_to_not_convert):
                in_features = module.in_features
                out_features = module.out_features

                model._modules[name] = target_cls(
                    w_bit=quantization_config.bits,
                    group_size=quantization_config.group_size,
                    in_features=in_features,
                    out_features=out_features,
                    bias=module.bias is not None,
                    dev=module.weight.device,
                )
                has_been_replaced = True

                # Force requires grad to False to avoid unexpected errors
                model._modules[name].requires_grad_(False)
        if len(list(module.children())) > 0:
            _, has_been_replaced = replace_with_awq_linear(
                module,
                modules_to_not_convert=modules_to_not_convert,
                current_key_name=current_key_name,
                quantization_config=quantization_config,
                has_been_replaced=has_been_replaced,
            )
        # Remove the last key for recursion
        current_key_name.pop(-1)
    return model, has_been_replaced


def get_modules_to_fuse(model, quantization_config):
    """
    Returns the fusing mapping given the quantization config and the model

    Args:
        model (`~PreTrainedModel`):
            The model to fuse - note this model should have been converted into AWQ format beforehand.
        quantization_config (`~transformers.quantization_config.AWQConfig`):
            The quantization configuration to use.
    """
    if not isinstance(model, PreTrainedModel):
        raise ValueError(f"The model should be an instance of `PreTrainedModel`, got {model.__class__.__name__}")

    # Always default to `quantization_config.modules_to_fuse`
    if quantization_config.modules_to_fuse is not None:
        current_fused_mapping = quantization_config.modules_to_fuse
        current_fused_mapping["max_seq_len"] = quantization_config.fuse_max_seq_len
    elif model.config.model_type in AWQ_FUSED_MAPPINGS:
        current_fused_mapping = AWQ_FUSED_MAPPINGS[model.config.model_type]

        # Properly deal with the case where we have a multi-modal model as well (e.g. Llava)
        if not hasattr(model.config, "text_config"):
            config = model.config
        else:
            config = model.config.text_config

        # Handle hidden_size, num_attention_heads, num_key_value_heads on our own.
        hidden_size = config.hidden_size
        num_attention_heads = config.num_attention_heads
        num_key_value_heads = getattr(config, "num_key_value_heads", num_attention_heads)

        # Fill `current_fused_mapping` with the expected values
        current_fused_mapping["hidden_size"] = hidden_size
        current_fused_mapping["num_attention_heads"] = num_attention_heads
        current_fused_mapping["num_key_value_heads"] = num_key_value_heads
        current_fused_mapping["max_seq_len"] = quantization_config.fuse_max_seq_len
    else:
        raise ValueError(
            "Fusing mapping not found either on the quantization config or the supported `AWQ_FUSED_MAPPINGS`. Please pass a `fused_mapping` argument"
            " in the `quantization_config` or raise an issue on transformers https://github.com/huggingface/transformers to add its support."
        )
    return current_fused_mapping


def fuse_awq_modules(model, quantization_config):
    """
    Optionally fuse some modules in the model to speedup inference.

    Args:
        model (`~PreTrainedModel`):
            The model to fuse - note this model should have been converted into AWQ format beforehand.
        quantization_config (`Union[AwqConfig, dict]`):
            The quantization configuration to use.
    """
    # We need to convert it from dict in order to get an AwqConfig object
    # otherwise the fields `backend` etc. will not be available
    # https://github.com/huggingface/transformers/pull/27411#discussion_r1414044495
    if isinstance(quantization_config, dict):
        quantization_config = AwqConfig.from_dict(quantization_config)
    backend = quantization_config.backend

    modules_to_fuse = get_modules_to_fuse(model, quantization_config)
    modules_to_not_convert = getattr(quantization_config, "modules_to_not_convert", None)

    if backend == AwqBackendPackingMethod.AUTOAWQ:
        from awq.modules.fused.attn import QuantAttentionFused
        from awq.modules.fused.mlp import QuantFusedMLP
        from awq.modules.fused.norm import FasterTransformerRMSNorm
    else:
        raise ValueError("Fusing is only supported for the AutoAWQ backend")

    fused_attention_modules = []

    for name, module in model.named_modules():
        if modules_to_not_convert is not None:
            if any(module_name_to_not_convert in name for module_name_to_not_convert in modules_to_not_convert):
                continue

        # Replace layer norms
        _fuse_awq_layernorm(modules_to_fuse["layernorm"], module, FasterTransformerRMSNorm)

        # Replace MLP layers
        _fuse_awq_mlp(model, name, modules_to_fuse["mlp"], module, QuantFusedMLP)

        # Replace attention layers
        attention_has_been_fused = _fuse_awq_attention_layers(
            model, module, modules_to_fuse, name, QuantAttentionFused
        )

        if attention_has_been_fused:
            fused_attention_modules.append(name.split(".")[0])

    # For AWQ fused + Llama we need to set `config._attn_implementation` = "custom" to avoid unexpected behavior and pass
    # `None` attention mask to the fused attention modules as now the attention mask is dropped by our models and dealt
    # by the `AttentionMaskConverter` module.
    if len(fused_attention_modules) > 0:
        for module_name, module in model.named_modules():
            if any(
                module_name in fused_attention_modules for fused_attention_parent_module in fused_attention_modules
            ):
                if hasattr(module, "config") and hasattr(module.config, "_attn_implementation"):
                    module.config._attn_implementation = "custom"
    return model


def _fuse_awq_layernorm(fuse_module_names, module, target_cls):
    """
    Fuse the LayerNorm layers into a target class using autoawq

    Args:
        fuse_module_names (`List[str]`):
            The list of module names to fuse
        module (`nn.Module`):
            The pytorch parent module that has layernorm modules to fuse
        target_cls (`~autoawq.FasterTransformerRMSNorm`):
            The `FasterTransformerRMSNorm` class as it only supports that class
            for now.
    """
    for module_name in fuse_module_names:
        if hasattr(module, module_name):
            old_module = getattr(module, module_name)
            module._modules[module_name] = target_cls(
                old_module.weight,
                old_module.variance_epsilon,
            ).to(old_module.weight.device)
            del old_module


def _fuse_awq_mlp(model, current_module_name, fuse_module_names, module, target_cls):
    """
    Fuse the MLP layers into a target class using autoawq

    Args:
        model (`~PreTrainedModel`):
            The input pretrained model
        current_module_name (`str`):
            The current submodule name
        fuse_module_names (`List[str]`):
            The list of module names to fuse. For the MLP layers it has to be an array
            of length 3 that consists of the 3 MLP layers in the order (gate (dense layer post-attention) / up / down layers)
        module (`nn.Module`):
            The pytorch parent module that has layernorm modules to fuse
        target_cls (`~autoawq.QuantFusedMLP`):
            The `QuantFusedMLP` class as it only supports that class
            for now.
    """
    if len(fuse_module_names) == 0:
        return

    if hasattr(module, fuse_module_names[0]):
        gate_proj = getattr(module, fuse_module_names[0])
        up_proj = getattr(module, fuse_module_names[1])
        down_proj = getattr(module, fuse_module_names[2])

        previous_device = gate_proj.qweight.device

        # Deal also with the case model has `text_config` attribute
        hidden_act = (
            model.config.hidden_act
            if not hasattr(model.config, "text_config")
            else model.config.text_config.hidden_act
        )
        activation_fn = ACT2FN[hidden_act]
        new_module = target_cls(gate_proj, down_proj, up_proj, activation_fn)

        parent_name, child_name = current_module_name.rsplit(".", 1)
        parent = model.get_submodule(parent_name)
        setattr(parent, child_name, new_module.to(previous_device))

        del gate_proj, up_proj, down_proj


def _fuse_awq_attention_layers(model, module, modules_to_fuse, current_module_name, target_cls):
    """
    Fuse the Attention layers into a target class using autoawq

    Args:
        model (`~PreTrainedModel`):
            The input pretrained model
        module (`nn.Module`):
            The pytorch parent module that has layernorm modules to fuse
        modules_to_fuse (`List[str]`):
            The module fusing mapping. The dictionary has to contain a field `attention` with attention module names
            in the correct order: q, k, v, o layer
        current_module_name (`str`):
            The current submodule name
        target_cls (`~autoawq.QuantAttentionFused`):
            The `QuantAttentionFused` class as it only supports that class
            for now.
    """
    from awq.modules.linear import WQLinear_GEMM, WQLinear_GEMV

    module_has_been_fused = False

    if len(modules_to_fuse["attention"]) == 0:
        return module_has_been_fused

    if hasattr(module, modules_to_fuse["attention"][0]):
        # First, we pack the QKV layers together
        q_proj = getattr(module, modules_to_fuse["attention"][0])

        if isinstance(q_proj, WQLinear_GEMV):
            linear_target_cls = WQLinear_GEMV
            cat_dim = 0
        elif isinstance(q_proj, WQLinear_GEMM):
            linear_target_cls = WQLinear_GEMM
            cat_dim = 1
        else:
            raise ValueError("Unsupported q_proj type: {type(q_proj)}")

        previous_device = q_proj.qweight.device

        k_proj = getattr(module, modules_to_fuse["attention"][1])
        v_proj = getattr(module, modules_to_fuse["attention"][2])
        o_proj = getattr(module, modules_to_fuse["attention"][3])

        bias = torch.cat([q_proj.bias, k_proj.bias, v_proj.bias], dim=0) if q_proj.bias is not None else None

        qkv_layer = linear_target_cls(
            q_proj.w_bit,
            q_proj.group_size,
            q_proj.in_features,
            q_proj.out_features + k_proj.out_features + v_proj.out_features,
            q_proj.bias is not None,
            next(iter(module.state_dict().values())).device,
        )

        qkv_layer.qweight = torch.cat([q_proj.qweight, k_proj.qweight, v_proj.qweight], dim=cat_dim)
        qkv_layer.qzeros = torch.cat([q_proj.qzeros, k_proj.qzeros, v_proj.qzeros], dim=cat_dim)
        qkv_layer.scales = torch.cat([q_proj.scales, k_proj.scales, v_proj.scales], dim=cat_dim)

        if isinstance(qkv_layer, WQLinear_GEMV):
            qkv_layer.split_k_iters = q_proj.split_k_iters

        qkv_layer.bias = bias

        fused_attention_layer = target_cls(
            modules_to_fuse["hidden_size"],
            modules_to_fuse["num_attention_heads"],
            modules_to_fuse["num_key_value_heads"],
            qkv_layer,
            o_proj,
            previous_device,
            modules_to_fuse["max_seq_len"],
            use_alibi=modules_to_fuse["use_alibi"],
            # The default value in autoawq is set to 10000.0
            rope_theta=modules_to_fuse.get("rope_theta", 10000.0),
        )

        fused_attention_layer.is_hf_transformers = True

        parent_name, child_name = current_module_name.rsplit(".", 1)
        parent = model.get_submodule(parent_name)
        setattr(parent, child_name, fused_attention_layer.to(previous_device))

        del q_proj, k_proj, v_proj, o_proj
        module_has_been_fused = True

    return module_has_been_fused


def post_init_awq_exllama_modules(model, exllama_config):
    """
    Runs post init for Exllama layers which performs:
        - Weights unpacking, reordering and repacking
        - Devices scratch space allocation
    """

    if exllama_config["version"] == ExllamaVersion.ONE:
        from awq.modules.linear.exllama import exllama_post_init

        model = exllama_post_init(model)
    elif exllama_config["version"] == ExllamaVersion.TWO:
        from awq.modules.linear.exllamav2 import exllamav2_post_init

        model = exllamav2_post_init(
            model,
            max_input_len=exllama_config["max_input_len"],
            max_batch_size=exllama_config["max_batch_size"],
        )
    else:
        raise ValueError(f"Unrecognized Exllama version: {exllama_config['version']}")

    return model