modeling_ibs2.py

# coding=utf-8
# Copyright 2024 state-spaces/mamba2 org and HuggingFace Inc. team.
#
# 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.
"""PyTorch MAMBA2 model."""

import math
from dataclasses import dataclass
from typing import Optional, Tuple, Union

import torch
import torch.utils.checkpoint
from torch import nn
from torch.nn import CrossEntropyLoss

from transformers.activations import ACT2FN
from transformers.generation import GenerationMixin
from transformers.modeling_utils import PreTrainedModel
from transformers.utils import (
    ModelOutput,
    add_code_sample_docstrings,
    add_start_docstrings,
    add_start_docstrings_to_model_forward,
    logging,
)
from transformers.utils.import_utils import is_causal_conv1d_available, is_torch_available, _is_package_available, version
from .configuration_ibs2 import IBS2Config

def is_mamba_2_ssm_available():
    if is_torch_available():
        import torch

        if not torch.cuda.is_available():
            return False
        else:
            if _is_package_available("mamba_ssm"):
                import mamba_ssm

                if version.parse(mamba_ssm.__version__) >= version.parse("2.0.4"):
                    return True
    return False

logger = logging.get_logger(__name__)


if is_mamba_2_ssm_available():
    from mamba_ssm.ops.triton.selective_state_update import selective_state_update
    from mamba_ssm.ops.triton.ssd_combined import mamba_chunk_scan_combined, mamba_split_conv1d_scan_combined
else:
    mamba_chunk_scan_combined, mamba_split_conv1d_scan_combined, selective_state_update = None, None, None

if is_causal_conv1d_available():
    from causal_conv1d import causal_conv1d_fn, causal_conv1d_update
else:
    causal_conv1d_update, causal_conv1d_fn = None, None

is_fast_path_available = all(
    (
        selective_state_update,
        mamba_chunk_scan_combined,
        mamba_split_conv1d_scan_combined,
        causal_conv1d_fn,
        causal_conv1d_update,
    )
)

_CHECKPOINT_FOR_DOC = "mistralai/mamba-codestral-7B-v0.1"
_CONFIG_FOR_DOC = "Mamba2Config"


# Helper methods for segment sum computation


def pad_tensor_by_size(input_tensor: torch.Tensor, pad_size: int):
    """
    Padding x tensor with `pad_size` on the seq_len dim (dim=1)

    Assumes that we only have tensors of either size 4 or 3
    """
    pad_shape = (0, 0, 0, 0, 0, pad_size, 0, 0) if len(input_tensor.shape) == 4 else (0, 0, 0, pad_size, 0, 0)

    return torch.nn.functional.pad(input_tensor, pad_shape, mode="constant", value=0)


def reshape_into_chunks(input_tensor, pad_size, chunk_size):
    """
    Padding input_tensor with `pad_size` on the seq_len dim (dim=1) and
    simultaneously splitting it into chunk sequences.

    Assumes that we only have tensors of either size 4 or 3
    """
    # [bsz, seq_len, ...] -> [bsz, seq_len multiple of chunk_size, ...]
    input_tensor = pad_tensor_by_size(input_tensor, pad_size)

    if len(input_tensor.shape) == 3:
        # [bsz, seq_len multiple of chunk_size, num_heads] -> [bsz, -1, chunk_size, num_heads]
        return input_tensor.reshape(input_tensor.shape[0], -1, chunk_size, input_tensor.shape[2])
    else:
        # [bsz, seq_len multiple of chunk_size, num_heads, head_dim or state_size] -> [bsz, -1, chunk_size, num_heads, head_dim or state_size]
        return input_tensor.reshape(
            input_tensor.shape[0], -1, chunk_size, input_tensor.shape[2], input_tensor.shape[3]
        )


def segment_sum(input_tensor):
    """
    More stable segment sum calculation. Uses cumulative sums and masking instead of direct subtractions.
    """
    chunk_size = input_tensor.size(-1)
    # 1. expand input tensor to have an additional dimension and repeat along that dimension
    # [..., chunk_size] -> [..., chunk_size, chunk_size]
    input_tensor = input_tensor[..., None].expand(*input_tensor.size(), chunk_size)
    # 2. create a lower triangular mask with the diagonal set to 0 to 0 out elements above diag
    mask = torch.tril(torch.ones(chunk_size, chunk_size, device=input_tensor.device, dtype=torch.bool), diagonal=-1)
    input_tensor = input_tensor.masked_fill(~mask, 0)
    # 3. compute actual cumsum
    tensor_segsum = torch.cumsum(input_tensor, dim=-2)

    # 4. apply mask to keep only the lower triangular part of the cumulative sum result (incl diagonal this time)
    mask = torch.tril(torch.ones(chunk_size, chunk_size, device=input_tensor.device, dtype=torch.bool), diagonal=0)
    tensor_segsum = tensor_segsum.masked_fill(~mask, -torch.inf)
    return tensor_segsum


def apply_mask_to_padding_states(hidden_states, attention_mask):
    """
    Tunes out the hidden states for padding tokens, see https://github.com/state-spaces/mamba/issues/66
    """
    if attention_mask is not None and attention_mask.shape[1] > 1 and attention_mask.shape[0] > 1:
        dtype = hidden_states.dtype
        hidden_states = (hidden_states * attention_mask[:, :, None]).to(dtype)

    return hidden_states


class Mamba2Cache:
    """
    Arguments:
        config: Mamba2Config
        batch_size: int
        dtype: torch.dtype
        device: torch.device

    Attributes:
        dtype: (`torch.dtype`):
            The default `dtype` used to initializing the cache.
        conv_kernel_size: (`int`):
            Model's convolution kernel size taken from config.
        n_groups: (`int`):
            Model's number of groups taken from the config - similar to tensor parallel in Transformer.
        state_size: (`int`):
            Model's SSM state size taken from config.
        num_heads: (`int`):
            The number of heads used in the linear attention / SSM.
        head_dim: (`int`):
            The respective dimension of the heads used in the linear attention / SSM.
        intermediate_size: (`int`):
            Model's intermediate_size based on (expand * hidden_dim) from config.
        conv_states: (`torch.Tensor`):
            A tensor of shape `[num_layers, batch_size, conv_kernel_size, intermediate_size + 2 * n_groups * state_size]` that holds convolutional states.
        ssm_states: (`torch.Tensor`):
            A tensor of shape `[num_layers, batch_size, num_heads, head_dim, state_size]` that holds ssm states.
    """

    def __init__(
        self, config: IBS2Config, batch_size: int, dtype: torch.dtype = torch.float16, device: Optional[str] = None
    ):
        self.dtype = dtype
        self.conv_kernel_size = config.conv_kernel
        self.n_groups = config.n_groups
        self.state_size = config.state_size
        self.num_heads = config.num_heads
        self.head_dim = config.head_dim
        self.intermediate_size = int(config.expand * config.hidden_size)

        self.conv_states = torch.zeros(
            config.num_hidden_layers,
            batch_size,
            self.intermediate_size + 2 * self.n_groups * self.state_size,
            self.conv_kernel_size,
            device=device,
            dtype=dtype,
        )
        self.ssm_states = torch.zeros(
            config.num_hidden_layers,
            batch_size,
            self.num_heads,
            self.head_dim,
            self.state_size,
            device=device,
            dtype=dtype,
        )

    def update_conv_state(
        self, layer_idx: int, new_conv_state: torch.Tensor, cache_init: bool = False
    ) -> torch.Tensor:
        if cache_init:
            self.conv_states[layer_idx] = new_conv_state.to(self.conv_states.device)
        else:
            self.conv_states[layer_idx] = self.conv_states[layer_idx].roll(shifts=-1, dims=-1)
            self.conv_states[layer_idx][:, :, -1] = new_conv_state[:, 0, :].to(self.conv_states.device)
        return self.conv_states[layer_idx]

    def update_ssm_state(self, layer_idx: int, new_ssm_state: torch.Tensor):
        self.ssm_states[layer_idx] = new_ssm_state.to(self.ssm_states.device)
        return self.ssm_states[layer_idx]

    def reset(self):
        self.conv_states.zero_()
        self.ssm_states.zero_()


class MambaRMSNormGated(torch.nn.Module):
    def __init__(self, hidden_size, eps=1e-6):
        super().__init__()
        self.weight = nn.Parameter(torch.ones(hidden_size))
        self.variance_epsilon = eps

    def forward(self, hidden_states, gate=None):
        input_dtype = hidden_states.dtype
        hidden_states = hidden_states.to(torch.float32)

        if gate is not None:
            hidden_states = hidden_states * nn.functional.silu(gate.to(torch.float32))
        variance = hidden_states.pow(2).mean(-1, keepdim=True)
        hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)

        return self.weight * hidden_states.to(input_dtype)

class Normalize(nn.Module):
    def __init__(self, min_value=None, max_value=None):
        super().__init__()
        self.min_value = min_value
        self.max_value = max_value

    def forward(self, value_states):
        # 计算 value_states 的最小值和最大值
        min_val = value_states.min(dim=-1, keepdim=True).values
        max_val = value_states.max(dim=-1, keepdim=True).values
        
        # 如果 min_value 或 max_value 是 None，则使用另一个值进行归一化
        if self.min_value is None and self.max_value is not None:
            # 只使用 max_value 进行归一化
            scale_factor = self.max_value / (max_val + 1e-6)
            return value_states * scale_factor, scale_factor, None
        elif self.max_value is None and self.min_value is not None:
            # 只使用 min_value 进行归一化
            scale_factor = self.min_value / (min_val + 1e-6)
            return value_states * scale_factor, scale_factor, None
        elif self.min_value is not None and self.max_value is not None:
            # 同时使用 min_value 和 max_value 进行归一化
            scale_factor = (self.max_value - self.min_value) / (max_val - min_val + 1e-6)
            shift_factor = self.min_value - min_val * scale_factor
            normalized_value_states = value_states * scale_factor + shift_factor
            return normalized_value_states, scale_factor, shift_factor
        else:
            # 如果 min_value 和 max_value 都是 None，则不进行归一化
            return value_states, None, None


# torch.lgamma(n) - torch.lgamma(theta) + (theta - n) * torch.special.digamma(theta)
class GammaIB(nn.Module):
    def __init__(self, hidden_size, alphas=None, return_attn=False, **kwargs) -> None:
        super().__init__()
        self.alphas = alphas
        # self.attributor = nn.Linear(hidden_size, 1)
        self.hidden_size = hidden_size
        self._auxiliary_loss = 0
        self.epoch_frac = 0
        self.epoch_threshold = -1
        self.normalizer = Normalize(max_value=10, min_value=0.1)
        self.return_attn = return_attn
        self._attn = None



    def get_auxiliary_loss(self):
        loss = self._auxiliary_loss
        self._auxiliary_loss = 0.0
        # print("auxiliary_loss", loss)
        return loss

    def init_alphas(self, param_alphas):
        # shape: [bsz, seq_len, 1]
        if self.alphas is None:
            maxmimum = 8
        else:
            maxmimum = param_alphas.size(1) / self.alphas.size(0) * self.alphas.max().item()
        length = param_alphas.shape[1]
        alphas = torch.linspace(maxmimum, 1, steps=length).float().to(param_alphas.device) # distance-decay: torch.linspace(1, 2, steps=length)
        self.alphas = alphas

    def compute_loss(self, param_alphas, epsilon=1e-6):
        if self.alphas is None:
            self.init_alphas(param_alphas)  # length is the second dimension of att
            print(f"Gamma prior alpha first: {self.alphas[0]}, last: {self.alphas[-1]}, size: {self.alphas.size(0)}", )
        if self.alphas.size(0) != param_alphas.size(1):
            self.init_alphas(param_alphas)  # length is the second dimension of att
            print(f"Gamma prior alpha first: {self.alphas[0]}, last: {self.alphas[-1]}, size: {self.alphas.size(0)}", )
        
        params = self.alphas.unsqueeze(-1).expand(param_alphas.shape)
        reg_loss = (torch.lgamma(self.alphas) - torch.lgamma(params) + (params - self.alphas) * torch.digamma(params)).mean()
        
        return reg_loss

    def forward(self, states):
        # hidden_states shape [bsz, seq_length, dimension]
        hidden_states, alphas = torch.split(
                    states,
                    [self.hidden_size, 1],
                    dim=-1
                )
        if self.epoch_frac < self.epoch_threshold:
            return hidden_states
        # alphas = F.rms_norm(alphas, [hidden_states.shape[1], 1])
        #! HACK: for finetuning we zero-init and reparametrize to ensure the initialization is one
        alphas = nn.functional.softplus(alphas) - torch.log(torch.tensor(2.0)) + torch.tensor(1.0)
        # if self.return_attn:
        #     # print(alphas.shape)
        #     if alphas.shape[-2] == 1:
        #         pass # seqlen == 1 indicates it is caching
        #     else:
        #         self._attn = alphas.detach().cpu()

        shaped_alphas = alphas.expand(-1, -1, hidden_states.shape[2])
        if self.training:
            self._auxiliary_loss = self.compute_loss(alphas)
            value_states = hidden_states.abs()  # ensure value_states is positive
            normalized_value_states, scale_factor, shift_factor = self.normalizer(value_states)

            betas = torch.reciprocal(normalized_value_states)
            sign_states = torch.sign(hidden_states)
            gamma_dist = torch.distributions.gamma.Gamma(shaped_alphas, betas)
            samples = gamma_dist.rsample()
            # Restore the original scale
            if shift_factor is not None:
                time_states = samples * sign_states / scale_factor - shift_factor / scale_factor
            else:
                time_states = samples * sign_states / scale_factor
        else:
            time_states = shaped_alphas * hidden_states # directly use the expectation
        
        if self.return_attn:
            # print(alphas.shape)
            if time_states.shape[-2] == 1:
                pass # seqlen == 1 indicates it is caching
            else:
                self._attn = torch.exp(-time_states).mean(dim=-1).detach().cpu()

        return time_states

class BernoulliIB(nn.Module):
    def __init__(self, hidden_size, temp=1, thetas=None, max_seqlen=1024, return_attn=False, **kwargs) -> None:
        super().__init__()
        self.epoch_frac = 0
        self.epoch_threshold = 0
        self.temp = temp
        self.thetas = thetas
        self.hidden_size = hidden_size
        # self.attributor = nn.Linear(hidden_size, 1, bias=True)
        self._auxiliary_loss = 0
        self.max_seqlen = 4096
        self.return_attn = return_attn
        self._attn = None

    def init_thetas(self, attn):
        # Create a tensor with sequence positions [0, 1, ..., length-1]
        # length = attn.shape[1]
        seq_len = attn.shape[1]
        if seq_len <= self.max_seqlen:
            positions = torch.arange(self.max_seqlen).float().to(attn.device)
        else:
            positions = torch.arange(self.max_seqlen - seq_len, self.max_seqlen).float().to(attn.device)
        
        # Define the exponential decay function
        # decay_factor = 0.3 + 0.4 * torch.exp(positions / length - 1) # extrapolable distance-decay (-\infty: 0.3, 0: 0.5, length: 0.7)
        decay_factor = 0.8 - 0.6 * torch.exp(positions / self.max_seqlen - 1) # extrapolable distance-balance (-\infty: 0.7, 0: 0.5, length: 0.3)
        
        # Make the decay factor repeat across the batch dimension
        self.thetas = decay_factor
        return decay_factor

    # def get_token_saliency(self):
    #     attn = self._attn
    #     self._attn = None
    #     return attn

    def get_auxiliary_loss(self):
        loss = self._auxiliary_loss
        self._auxiliary_loss = 0.0
        return loss

    def compute_loss(self, att, epsilon=1e-6):
        if self.thetas is None:
            thetas = self.init_thetas(att)  # length is the second dimension of att
            print(f"Bernoulli prior theta first: {self.thetas[0]:.2f}, last: {self.thetas[-1]:.2f}, size: {self.thetas.size(0)}")

        if self.thetas.size(0) >= att.size(1):
            thetas = self.thetas[-att.size(1):]
        elif self.thetas.size(0) < att.size(1):
            thetas = self.init_thetas(att)
            print(f"Bernoulli prior theta first: {self.thetas[0]:.2f}, last: {self.thetas[-1]:.2f}, size: {self.thetas.size(0)}")
        
        thetas = thetas.unsqueeze(-1).expand(att.shape)
        # Calculate the regularization loss
        reg_loss = (att * torch.log(att / thetas + epsilon) +
                    (1 - att) * torch.log((1 - att) / (1 - thetas + epsilon) + epsilon)).mean()
        
        return reg_loss

    def forward(self, states):
        # hidden_states shape [bsz, seq_length, dimension]
        hidden_states, attn = torch.split(
                    states,
                    [self.hidden_size, 1],
                    dim=-1
                )
        # attn = self.attributor(hidden_states)
        #! HACK: for fintuning, we zero-init and re-paramerize as plus 1
        attn = attn + torch.tensor(1.0)
        if self.epoch_frac < self.epoch_threshold:
            return hidden_states
        if self.training:
            # gumble soft-max
            random_noise = torch.empty_like(attn).uniform_(1e-10, 1 - 1e-10)
            random_noise = torch.log(random_noise) - torch.log(1.0 - random_noise)
            attn_bern = ((attn + random_noise) / self.temp).sigmoid()
        else:
            attn_bern = (attn).sigmoid()
        self._auxiliary_loss = self.compute_loss(attn_bern)
        if self.return_attn:
            if attn_bern.shape[-2] == 1:
                pass # seqlen == 1 indicates it is caching
            else:
                self._attn = attn_bern.detach().cpu()
        return hidden_states * attn_bern


class Mamba2Mixer(nn.Module):
    """
    Compute ∆, A, B, C, and D the state space parameters and compute the `contextualized_states`.
    A, D are input independent (see Mamba paper [1] Section 3.5.2 "Interpretation of A" for why A isn't selective)
    ∆, B, C are input-dependent (this is a key difference between Mamba and the linear time invariant S4,
    and is why Mamba is called **selective** state spaces)
    """

    def __init__(self, config: IBS2Config, layer_idx: int):
        super().__init__()
        self.num_heads = config.num_heads
        self.hidden_size = config.hidden_size
        self.ssm_state_size = config.state_size
        self.conv_kernel_size = config.conv_kernel
        self.intermediate_size = int(config.expand * self.hidden_size)
        self.time_step_rank = int(config.time_step_rank)
        self.layer_idx = layer_idx
        self.use_conv_bias = config.use_conv_bias
        self.activation = config.hidden_act
        self.act = ACT2FN[config.hidden_act]

        self.layer_norm_epsilon = config.layer_norm_epsilon
        self.rms_norm = config.rms_norm

        self.n_groups = config.n_groups
        self.head_dim = config.head_dim
        self.chunk_size = config.chunk_size

        self.time_step_limit = config.time_step_limit
        self.time_step_min = config.time_step_min
        self.time_step_max = config.time_step_max

        self.conv_dim = self.intermediate_size + 2 * self.n_groups * self.ssm_state_size
        self.conv1d = nn.Conv1d(
            in_channels=self.conv_dim,
            out_channels=self.conv_dim,
            bias=config.use_conv_bias,
            kernel_size=config.conv_kernel,
            groups=self.conv_dim,
            padding=config.conv_kernel - 1,
        )

        # projection of the input hidden states
        #! HACK ib_dim
        self._attn = None
        self.return_attn = config.return_attn
        assert config.ib_type in ['bernoulli', 'gamma'], "Invalid IB Prior."
        IB_cls = BernoulliIB if config.ib_type == 'bernoulli' else GammaIB if config.ib_type == 'gamma' else None
        self.ib4dt = IB_cls(self.num_heads, return_attn=config.return_attn) if self.layer_idx in [0, 31, 63] else None

        self.ib_proj = nn.Linear(
            self.hidden_size,
            1,
            bias=False,
        ) if self.ib4dt else None
        projection_size = self.intermediate_size + self.conv_dim + self.num_heads
        self.in_proj = nn.Linear(
            self.hidden_size,
            projection_size,
            bias=config.use_bias,
        )
        # selective projection used to make dt, B and C input dependant

        # time step projection (discretization)
        # instantiate once and copy inv_dt in init_weights of PretrainedModel
        self.dt_bias = nn.Parameter(torch.ones(self.num_heads))

        # S4D real initialization. These are not discretized!
        # The core is to load them, compute the discrete states, then write the updated state. Keeps the memory bounded
        A = torch.arange(1, self.num_heads + 1)
        self.A_log = nn.Parameter(torch.log(A))
        self.A_log._no_weight_decay = True
        self.norm = MambaRMSNormGated(self.intermediate_size, eps=self.layer_norm_epsilon)
        self.D = nn.Parameter(torch.ones(self.num_heads))
        self.D._no_weight_decay = True

        self.out_proj = nn.Linear(self.intermediate_size, self.hidden_size, bias=config.use_bias)
        self.use_bias = config.use_bias

        if not is_fast_path_available:
            logger.warning_once(
                "The fast path is not available because on of `(selective_state_update, causal_conv1d_fn, causal_conv1d_update)`"
                " is None. Falling back to the naive implementation. To install follow https://github.com/state-spaces/mamba/#installation and"
                " https://github.com/Dao-AILab/causal-conv1d"
            )

    def get_token_saliency(self):
        attn = self._attn
        self._attn = None
        return attn

    def cuda_kernels_forward(
        self,
        hidden_states: torch.Tensor,
        cache_params: Optional[Mamba2Cache] = None,
        cache_position: Optional[torch.LongTensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
    ):
        # 1. Gated MLP's linear projection
        hidden_states = apply_mask_to_padding_states(hidden_states, attention_mask)
        projected_states = self.in_proj(hidden_states)

        #! HACK IBS apply
        if self.ib4dt:
            ib_state = self.ib_proj(hidden_states)
            dim = self.head_dim * self.num_heads 
            zx, BC, dt = torch.split(projected_states, [dim * 2, + self.n_groups * self.ssm_state_size * 2, self.num_heads], dim=-1)
            dt = self.ib4dt(torch.cat([dt, ib_state], dim=-1))
            projected_states = torch.cat([zx, BC, dt], dim=-1)
            if self.return_attn and dt.shape[-2] != 1:
                dt_plus = nn.functional.softplus(dt + self.dt_bias)
                dA = (dt_plus * (-torch.exp(self.A_log.float())))
                # dA = torch.exp(dA)
                attn = dA.mean(dim=-1) # - 0.1 * dA.std(dim=-1) # attn shape [batch_size, seqlen]
                self._attn = attn

        # Set up dimensions for reshapes later
        batch_size, seq_len, _ = hidden_states.shape
        groups_time_state_size = self.n_groups * self.ssm_state_size
        d_mlp = (
            projected_states.shape[-1]
            - 2 * self.intermediate_size
            - 2 * self.n_groups * self.ssm_state_size
            - self.num_heads
        ) // 2

        # Single step calculations via cache
        if cache_params is not None and cache_position is not None and cache_position[0] > 0:
            _, _, gate, hidden_states_B_C, dt = projected_states.squeeze(1).split(
                [d_mlp, d_mlp, self.intermediate_size, self.conv_dim, self.num_heads], dim=-1
            )

            # 2. Convolution sequence transformation
            hidden_states_B_C = causal_conv1d_update(
                hidden_states_B_C,
                cache_params.conv_states[self.layer_idx],
                self.conv1d.weight.squeeze(1),
                self.conv1d.bias,
                self.activation,
            )

            hidden_states, B, C = torch.split(
                hidden_states_B_C,
                [self.intermediate_size, groups_time_state_size, groups_time_state_size],
                dim=-1,
            )

            # 3. SSM transformation
            A = -torch.exp(self.A_log.float())  # (nheads,)
            A = A[:, None, ...][:, :, None].expand(-1, self.head_dim, self.ssm_state_size).to(dtype=torch.float32)
            dt = dt[:, :, None].expand(-1, -1, self.head_dim)
            dt_bias = self.dt_bias[:, None, ...].expand(-1, self.head_dim)
            D = self.D[:, None, ...].expand(-1, self.head_dim)
            B = B.view(batch_size, self.n_groups, B.shape[1] // self.n_groups)
            C = C.view(batch_size, self.n_groups, C.shape[1] // self.n_groups)
            hidden_states_reshaped = hidden_states.view(batch_size, self.num_heads, self.head_dim)
            hidden_states = selective_state_update(
                cache_params.ssm_states[self.layer_idx],
                hidden_states_reshaped,
                dt,
                A,
                B,
                C,
                D,
                z=None,
                dt_bias=dt_bias,
                dt_softplus=True,
            )
            hidden_states = hidden_states.view(batch_size, self.num_heads * self.head_dim)
            hidden_states = self.norm(hidden_states, gate)

            # 4. Final linear projection
            out = self.out_proj(hidden_states)[:, None, ...]

        # Fused calculations or step by step if no initialized cache is found
        else:
            A = -torch.exp(self.A_log.float())  # (num_heads) or (intermediate_size, state_size)
            dt_limit_kwargs = {} if self.time_step_limit == (0.0, float("inf")) else {"dt_limit": self.time_step_limit}

            # 2-4. Fused kernel for conv1d, SSM, and the final projection
            if self.training and cache_params is None:
                out = mamba_split_conv1d_scan_combined(
                    projected_states,
                    self.conv1d.weight.squeeze(1),
                    self.conv1d.bias,
                    self.dt_bias,
                    A,
                    D=self.D,
                    chunk_size=self.chunk_size,
                    seq_idx=None,  # was seq_idx
                    activation=self.activation,
                    rmsnorm_weight=self.norm.weight,
                    rmsnorm_eps=self.norm.variance_epsilon,
                    outproj_weight=self.out_proj.weight,
                    outproj_bias=self.out_proj.bias,
                    headdim=self.head_dim,
                    ngroups=self.n_groups,
                    norm_before_gate=False,
                    return_final_states=False,
                    **dt_limit_kwargs,
                )

            else:
                _, _, gate, hidden_states_B_C, dt = projected_states.split(
                    [d_mlp, d_mlp, self.intermediate_size, self.conv_dim, self.num_heads], dim=-1
                )

                # 2. Convolution sequence transformation
                # Init cache
                if cache_params is not None:
                    hidden_states_B_C_transposed = hidden_states_B_C.transpose(1, 2)
                    conv_states = nn.functional.pad(
                        hidden_states_B_C_transposed,
                        (cache_params.conv_kernel_size - hidden_states_B_C_transposed.shape[-1], 0),
                    )
                    cache_params.update_conv_state(
                        layer_idx=self.layer_idx, new_conv_state=conv_states, cache_init=True
                    )

                if self.activation not in ["silu", "swish"]:
                    hidden_states_B_C = self.act(
                        self.conv1d(hidden_states_B_C.transpose(1, 2))[..., :seq_len].transpose(1, 2)
                    )
                else:
                    hidden_states_B_C = causal_conv1d_fn(
                        x=hidden_states_B_C.transpose(1, 2),
                        weight=self.conv1d.weight.squeeze(1),
                        bias=self.conv1d.bias,
                        activation=self.activation,
                    ).transpose(1, 2)

                hidden_states_B_C = apply_mask_to_padding_states(hidden_states_B_C, attention_mask)
                hidden_states, B, C = torch.split(
                    hidden_states_B_C,
                    [self.intermediate_size, groups_time_state_size, groups_time_state_size],
                    dim=-1,
                )

                # 3. SSM transformation
                scan_output, ssm_state = mamba_chunk_scan_combined(
                    hidden_states.view(batch_size, seq_len, -1, self.head_dim),
                    dt,
                    A,
                    B.view(batch_size, seq_len, self.n_groups, -1),
                    C.view(batch_size, seq_len, self.n_groups, -1),
                    chunk_size=self.chunk_size,
                    D=self.D,
                    z=None,
                    seq_idx=None,
                    return_final_states=True,
                    dt_bias=self.dt_bias,
                    dt_softplus=True,
                    **dt_limit_kwargs,
                )

                # Init cache
                if ssm_state is not None and cache_params is not None:
                    cache_params.update_ssm_state(layer_idx=self.layer_idx, new_ssm_state=ssm_state)

                scan_output = scan_output.view(batch_size, seq_len, -1)
                # Multiply "gate" branch and apply extra normalization layer
                scan_output = self.norm(scan_output, gate)

                # 4. Final linear projection
                out = self.out_proj(scan_output)
        return out

    # fmt: off
    def torch_forward(self, input_states, cache_params: Optional[Mamba2Cache]=None, cache_position:Optional[torch.LongTensor]=None, attention_mask: Optional[torch.Tensor]=None):
        batch_size, seq_len, _ = input_states.shape
        dtype = input_states.dtype

        # 1. Gated MLP's linear projection
        input_states = apply_mask_to_padding_states(input_states, attention_mask)
        projected_states = self.in_proj(input_states)
        
        if self.ib4dt:
            attn = self.ib_proj(input_states)
            dim = self.head_dim * self.num_heads 
            zx, BC, dt = torch.split(projected_states, [dim * 2, + self.n_groups * self.ssm_state_size * 2, self.num_heads], dim=-1)
            dt = self.ib4dt(torch.cat([dt, attn], dim=-1))
            projected_states = torch.cat([zx, BC, dt], dim=-1)

        d_mlp = (projected_states.shape[-1] - 2 * self.intermediate_size - 2 * self.n_groups * self.ssm_state_size-self.num_heads) // 2
        _, _, gate, hidden_states_B_C, dt = projected_states.split(
                [d_mlp, d_mlp, self.intermediate_size,  self.conv_dim, self.num_heads], dim=-1
        )

        # 2. Convolution sequence transformation
        if cache_params is not None and cache_position is not None and cache_position[0] > 0:
            cache_params.update_conv_state(layer_idx=self.layer_idx, new_conv_state=hidden_states_B_C, cache_init=False)

            # We need to guarantee that anything regarding the cache is on the same device
            conv_states = cache_params.conv_states[self.layer_idx].to(device=self.conv1d.weight.device)

            hidden_states_B_C = torch.sum(
                conv_states * self.conv1d.weight.squeeze(1), dim=-1
            )
            if self.use_conv_bias:
                hidden_states_B_C = hidden_states_B_C + self.conv1d.bias
            hidden_states_B_C = self.act(hidden_states_B_C)
        else:
            # Init cache
            if cache_params is not None:
                hidden_states_B_C_transposed = hidden_states_B_C.transpose(1, 2)
                conv_states = nn.functional.pad(
                    hidden_states_B_C_transposed, (cache_params.conv_kernel_size - hidden_states_B_C_transposed.shape[-1], 0)
                )
                cache_params.update_conv_state(layer_idx=self.layer_idx, new_conv_state=conv_states, cache_init=True)

            hidden_states_B_C = self.act(self.conv1d(hidden_states_B_C.transpose(1, 2))[..., :seq_len].transpose(1, 2))

        hidden_states_B_C = apply_mask_to_padding_states(hidden_states_B_C, attention_mask)
        hidden_states, B, C = torch.split(
            hidden_states_B_C,
            [self.intermediate_size, self.n_groups * self.ssm_state_size, self.n_groups * self.ssm_state_size],
            dim=-1
        )

        # 3. SSM transformation
        A = -torch.exp(self.A_log.float())                            # [num_heads]
        if cache_params is not None and cache_position is not None and cache_position[0] > 0:
            # We need to guarantee that anything regarding the cache is on the same device
            cache_device = cache_params.ssm_states.device

            # Note: there is no need to pad parameter matrices here, as there is just one new token
            # for batched generation
            dt = dt[:, 0, :][:, None, ...]
            dt = dt.transpose(1, 2).expand(batch_size, dt.shape[-1], self.head_dim)
            # [num_heads] -> [num_heads, head_dim]
            dt_bias = self.dt_bias[..., None].expand(self.dt_bias.shape[0], self.head_dim)

            dt = torch.nn.functional.softplus(dt + dt_bias.to(dt.dtype))
            dt = torch.clamp(dt, self.time_step_limit[0], self.time_step_limit[1])
            A = A[..., None, None].expand(self.num_heads, self.head_dim, self.ssm_state_size).to(dtype=torch.float32)
            # [bsz, num_heads, head_dim, state_size]
            dA = (torch.exp(dt[..., None] * A)).to(device=cache_device)

            # Discretize B
            # [bsz, n_groups * state_size] -> [bsz, n_groups, 1, state_size] ->
            # -> [bsz, n_groups, group to head repetition factor, state_size] -> [bsz, num_heads, state_size]
            B = B.reshape(batch_size, self.n_groups, -1)[..., None, :]
            B = B.expand(batch_size, self.n_groups, self.num_heads // self.n_groups, B.shape[-1]).contiguous()
            B = B.reshape(batch_size, -1, B.shape[-1])
            # [bsz, num_heads, head_dim, state_size]
            dB = dt[..., None] * B[..., None, :]

            # Discretize x into dB
            # [bsz, intermediate_size] -> [bsz, num_heads, head_dim]
            hidden_states = hidden_states.reshape(batch_size, -1, self.head_dim)
            dBx = (dB * hidden_states[..., None]).to(device=cache_device)

            # State calculation
            cache_params.update_ssm_state(
                layer_idx=self.layer_idx,
                new_ssm_state=cache_params.ssm_states[self.layer_idx] * dA + dBx
            )

            # Subsequent output
            # [bsz, n_groups * state_size] -> [bsz, num_heads, state_size]
            C = C.reshape(batch_size, self.n_groups, -1)[..., None, :]
            C = C.expand(batch_size, self.n_groups, self.num_heads // self.n_groups, C.shape[-1]).contiguous()
            C = C.reshape(batch_size, -1, C.shape[-1])
            # [bsz, num_heads, head_dim]

            ssm_states = cache_params.ssm_states[self.layer_idx].to(device=C.device, dtype=C.dtype)  # Shape: [b, h, d, n]
            # Reshape ssm_states to merge the first two dimensions
            ssm_states_reshaped = ssm_states.view(batch_size * self.num_heads, self.head_dim, self.ssm_state_size)  # Shape: [b*h, d, n]
            C_reshaped = C.view(batch_size * self.num_heads, self.ssm_state_size, 1)  # Shape: [b*h, n, 1]
            y = torch.bmm(ssm_states_reshaped, C_reshaped)
            y = y.view(batch_size, self.num_heads, self.head_dim)

            # D skip connection
            # [num_heads] -> [num_heads, head_dim]
            D = self.D[..., None].expand(self.D.shape[0], self.head_dim)
            y = (y + hidden_states * D).to(y.dtype)

            # [bsz, num_heads, head_dim] -> [bsz, 1, intermediate_size]
            y = y.reshape(batch_size, -1)[:, None, ...]
        else:
            # begin ssd naive implementation without einsums
            dt = nn.functional.softplus(dt + self.dt_bias)
            dt = torch.clamp(dt, self.time_step_limit[0], self.time_step_limit[1])
            hidden_states = hidden_states.reshape(batch_size, seq_len, -1, self.head_dim).float()
            B = B.reshape(batch_size, seq_len, -1, self.ssm_state_size).float()
            C = C.reshape(batch_size, seq_len, -1, self.ssm_state_size).float()
            B = B.repeat(1, 1, self.num_heads // self.n_groups, 1)
            C = C.repeat(1, 1, self.num_heads // self.n_groups, 1)
            pad_size = (self.chunk_size - seq_len % self.chunk_size) % self.chunk_size

            D_residual = self.D[..., None] * pad_tensor_by_size(hidden_states, pad_size)

            # Discretize x and A
            hidden_states = hidden_states * dt[..., None]
            A = A.to(hidden_states.dtype) * dt

            # Rearrange into blocks/chunks
            hidden_states, A, B, C = [reshape_into_chunks(t, pad_size, self.chunk_size) for t in (hidden_states, A, B, C)]

            # [bsz, -1, chunk_size, num_heads] -> [bsz, num_heads, -1, chunk_size]
            A = A.permute(0, 3, 1, 2)
            A_cumsum = torch.cumsum(A, dim=-1)

            # 1. Compute the output for each intra-chunk (diagonal blocks)
            # This is the analog of a causal mask
            L = torch.exp(segment_sum(A))

            # Contraction of C and B to get G (attention-weights like)
            G_intermediate = C[:, :, :, None, :, :] * B[:, :, None, :, :, :]  # shape: (b, c, l, s, h, n)
            G = G_intermediate.sum(dim=-1)  # shape: (b, c, l, s, h)

            # Compute M, equivalent to applying attention mask to weights
            M_intermediate = G[..., None] * L.permute(0, 2, 3, 4, 1)[..., None]
            M = M_intermediate.sum(dim=-1)

            # Compute Y_diag (apply to values)
            Y_diag = (M[..., None] * hidden_states[:, :, None]).sum(dim=3)

            # 2. Compute the state for each intra-chunk
            # (right term of low-rank factorization of off-diagonal blocks; B terms)
            decay_states = torch.exp((A_cumsum[:, :, :, -1:] - A_cumsum))
            B_decay = B * decay_states.permute(0, -2, -1, 1)[..., None]
            states = (B_decay[..., None, :] * hidden_states[..., None]).sum(dim=2)

            # 3. Compute the inter-chunk SSM recurrence; produces correct SSM states at chunk boundaries
            # (middle term of factorization of off-diag blocks; A terms)
            if cache_params is not None and cache_position is not None and cache_position[0] > 0:
                previous_states = cache_params.ssm_states[self.layer_idx][:, None, ...].to(device=states.device)
            else:
                previous_states = torch.zeros_like(states[:, :1])
            states = torch.cat([previous_states, states], dim=1)
            decay_chunk = torch.exp(segment_sum(nn.functional.pad(A_cumsum[:, :, :, -1], (1, 0))))
            decay_chunk = decay_chunk.transpose(1, 3)
            new_states = (decay_chunk[..., None, None] * states[:, :, None, ...]).sum(dim=1)
            states, ssm_state = new_states[:, :-1], new_states[:, -1]

            # 4. Compute state -> output conversion per chunk
            # (left term of low-rank factorization of off-diagonal blocks; C terms)
            state_decay_out = torch.exp(A_cumsum)
            C_times_states = (C[..., None, :] * states[:, :, None, ...])
            state_decay_out_permuted = state_decay_out.permute(0, 2, 3, 1)
            Y_off = (C_times_states.sum(-1) * state_decay_out_permuted[..., None])

            # Add output of intra-chunk and inter-chunk terms (diagonal and off-diagonal blocks)
            y = Y_diag + Y_off
            # [bsz, -1, self.chunk_size, num_heads, head_dim] -> [bsz, (padded) seq_len, num_heads, head_dim]
            y = y.reshape(batch_size, -1, self.num_heads, self.head_dim)

            y = y + D_residual
            # Cutting off padded chunks
            if pad_size > 0:
                y = y[:, :seq_len, :, :]
            y = y.reshape(batch_size, seq_len, -1)

            # Init cache
            if ssm_state is not None and cache_params is not None:
                cache_params.update_ssm_state(layer_idx=self.layer_idx, new_ssm_state=ssm_state)

        scan_output = self.norm(y, gate)

        # end ssd naive

        # 4. Final linear projection
        contextualized_states = self.out_proj(scan_output.to(dtype))  # [batch, seq_len, hidden_size]
        return contextualized_states
    # fmt: on

    def forward(
        self,
        hidden_states,
        cache_params: Optional[Mamba2Cache] = None,
        cache_position: Optional[torch.LongTensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
    ):
        if is_fast_path_available and "cuda" in self.in_proj.weight.device.type:
            return self.cuda_kernels_forward(hidden_states, cache_params, cache_position, attention_mask)
        dtype = hidden_states.dtype
        if attention_mask is not None and attention_mask.shape[1] > 1 and attention_mask.shape[0] > 1:
            # tune out hidden states for pad tokens, see https://github.com/state-spaces/mamba/issues/66
            hidden_states = (hidden_states * attention_mask[:, :, None]).to(dtype)

        return self.torch_forward(hidden_states, cache_params, cache_position, attention_mask)


class Mamba2RMSNorm(nn.Module):
    def __init__(self, hidden_size, eps=1e-6):
        """
        Mamba2RMSNorm is equivalent to T5LayerNorm and LlamaRMSNorm
        """
        super().__init__()
        self.weight = nn.Parameter(torch.ones(hidden_size))
        self.variance_epsilon = eps

    def forward(self, hidden_states):
        input_dtype = hidden_states.dtype
        hidden_states = hidden_states.to(torch.float32)
        variance = hidden_states.pow(2).mean(-1, keepdim=True)
        hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
        return self.weight * hidden_states.to(input_dtype)


class IBS2Block(nn.Module):
    def __init__(self, config, layer_idx):
        super().__init__()
        self.config = config
        self.layer_idx = layer_idx
        self.residual_in_fp32 = config.residual_in_fp32
        self.norm = Mamba2RMSNorm(config.hidden_size, eps=config.layer_norm_epsilon)
        self.mixer = Mamba2Mixer(config, layer_idx=layer_idx)

    def forward(
        self,
        hidden_states,
        cache_params: Optional[Mamba2Cache] = None,
        cache_position: Optional[torch.LongTensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
    ):
        residual = hidden_states
        hidden_states = self.norm(hidden_states.to(dtype=self.norm.weight.dtype))
        if self.residual_in_fp32:
            residual = residual.to(torch.float32)

        hidden_states = self.mixer(
            hidden_states, cache_params=cache_params, cache_position=cache_position, attention_mask=attention_mask
        )
        hidden_states = residual + hidden_states
        return hidden_states


class Mamba2PreTrainedModel(PreTrainedModel):
    """
    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
    models.
    """

    config_class = IBS2Config
    base_model_prefix = "backbone"
    _no_split_modules = ["Mamba2Block"]
    supports_gradient_checkpointing = True
    _is_stateful = True

    def _init_weights(self, module):
        """Initialize the weights."""
        if isinstance(module, Mamba2Mixer):
            #! HACK
            if getattr(module, "ib_proj"):
                nn.init.zeros_(module.ib_proj.weight)

            module.A_log._no_weight_decay = True
            module.D._no_weight_decay = True

            dt = torch.exp(
                torch.rand(self.config.num_heads)
                * (math.log(self.config.time_step_max) - math.log(self.config.time_step_min))
                + math.log(self.config.time_step_min)
            ).clamp(min=self.config.time_step_floor)

            # # Inverse of softplus: https://github.com/pytorch/pytorch/issues/72759
            inv_dt = dt + torch.log(-torch.expm1(-dt))
            with torch.no_grad():
                module.dt_bias.copy_(inv_dt)
            module.dt_bias._no_reinit = True

        if isinstance(module, nn.Linear):
            if module.bias is not None:
                if not getattr(module.bias, "_no_reinit", False):
                    nn.init.zeros_(module.bias)
        elif isinstance(module, nn.Embedding):
            nn.init.normal_(module.weight, std=self.config.initializer_range)

        if self.config.rescale_prenorm_residual:
            # Reinitialize selected weights subject to the OpenAI GPT-2 Paper Scheme:
            #   > A modified initialization which accounts for the accumulation on the residual path with model depth. Scale
            #   > the weights of residual layers at initialization by a factor of 1/√N where N is the # of residual layers.
            #   >   -- GPT-2 :: https://openai.com/blog/better-language-models/
            #
            # Reference (Megatron-LM): https://github.com/NVIDIA/Megatron-LM/blob/main/megatron/model/gpt_model.py
            for name, p in module.named_parameters():
                if name in ["out_proj.weight"]:
                    # Special Scaled Initialization --> There are 2 Layer Norms per Transformer Block
                    # Following Pytorch init, except scale by 1/sqrt(2 * n_layer)
                    # We need to reinit p since this code could be called multiple times
                    # Having just p *= scale would repeatedly scale it down
                    nn.init.kaiming_uniform_(p, a=math.sqrt(5))
                    with torch.no_grad():
                        p /= math.sqrt(self.config.num_hidden_layers)


@dataclass
# Copied from transformers.models.mamba.modeling_mamba.MambaOutput with MAMBA->MAMBA2,Mamba->Mamba2
class Mamba2Output(ModelOutput):
    """
    Class for the MAMBA2 model outputs.

    Args:
        last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
            Sequence of hidden-states at the output of the last layer of the model.
        cache_params (`Mamba2Cache`):
            The state of the model at the last time step. Can be used in a forward method with the next `input_ids` to
            avoid providing the old `input_ids`.

            Includes both the State space model state matrices after the selective scan, and the Convolutional states
        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
    """

    last_hidden_state: Optional[torch.FloatTensor] = None
    cache_params: Optional[Mamba2Cache] = None
    hidden_states: Optional[Tuple[torch.FloatTensor]] = None


@dataclass
# Copied from transformers.models.mamba.modeling_mamba.MambaCausalLMOutput with Mamba->Mamba2
class Mamba2CausalLMOutput(ModelOutput):
    """
    Base class for causal language model (or autoregressive) outputs.

    Args:
        loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Language modeling loss (for next-token prediction).
        logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
            Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
        cache_params (`Mamba2Cache`):
            The state of the model at the last time step. Can be used in a forward method with the next `input_ids` to
            avoid providing the old `input_ids`.

            Includes both the State space model state matrices after the selective scan, and the Convolutional states
        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
    """

    loss: Optional[torch.FloatTensor] = None
    logits: Optional[torch.FloatTensor] = None
    cache_params: Optional[Mamba2Cache] = None
    hidden_states: Optional[Tuple[torch.FloatTensor]] = None


MAMBA2_START_DOCSTRING = r"""

    This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
    library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
    etc.)

    This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
    Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
    and behavior.

    Parameters:
        config ([`Mamba2Config`]): Model configuration class with all the parameters of the model.
            Initializing with a config file does not load the weights associated with the model, only the
            configuration. Check out the [`~PreTrainedModel.from_pretrained`] method to load the model weights.
"""

MAMBA2_INPUTS_DOCSTRING = r"""
    Args:
        input_ids (`torch.LongTensor` of shape `(batch_size, input_ids_length)`):
            Indices of input sequence tokens in the vocabulary.

            If `cache_params.seqlen_offset>0`, only `input_ids` that do not have their past calculated should be passed as
            `input_ids`.

            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
            [`PreTrainedTokenizer.__call__`] for details.

            [What are input IDs?](../glossary#input-ids)
        inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
            Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
            is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
            model's internal embedding lookup matrix.
        cache_params (`Mamba2Cache`, *optional*):
            If passed along, the model uses the previous state in all the blocks (which will give the output for the
            `input_ids` provided as if the model add `state_input_ids + input_ids` as context).
        use_cache (`bool`, *optional*):
            If set to `True`, the `cache_params` is returned and can be used to quickly generate the next logits.
        output_hidden_states (`bool`, *optional*):
            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
            more detail.
        return_dict (`bool`, *optional*):
            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
        cache_position (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
            The position of the current input in the cache. This is used to ensure that the cache is correctly updated.
            If `cache_params` is passed, `cache_position` should also be passed.
        attention_mask (`torch.FloatTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:

            - 1 for tokens that are **not masked**,
            - 0 for tokens that are **masked**.

            [What are attention masks?](../glossary#attention-mask)
"""


@add_start_docstrings(
    "The bare MAMBA2 Model transformer outputting raw hidden-states without any specific head on top.",
    MAMBA2_START_DOCSTRING,
)
class IBS2Model(Mamba2PreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        self.embeddings = nn.Embedding(config.vocab_size, config.hidden_size)
        self.layers = nn.ModuleList([IBS2Block(config, layer_idx=idx) for idx in range(config.num_hidden_layers)])

        self.gradient_checkpointing = False
        self.norm_f = Mamba2RMSNorm(config.hidden_size, eps=config.layer_norm_epsilon)
        # Initialize weights and apply final processing
        self._register_load_state_dict_pre_hook(self.load_hook)
        self.post_init()

    def load_hook(self, state_dict, prefix, *args):
        for k in state_dict:
            if "embedding." in k:
                state_dict[k.replace("embedding.", "embeddings.")] = state_dict.pop(k)
                break

    def get_input_embeddings(self):
        return self.embeddings

    def set_input_embeddings(self, new_embeddings):
        self.embeddings = new_embeddings

    @add_start_docstrings_to_model_forward(MAMBA2_INPUTS_DOCSTRING)
    @add_code_sample_docstrings(
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=Mamba2Output,
        config_class=_CONFIG_FOR_DOC,
    )
    def forward(
        self,
        input_ids: Optional[torch.LongTensor] = None,
        inputs_embeds: Optional[torch.LongTensor] = None,
        cache_params: Optional[Mamba2Cache] = None,
        use_cache: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        cache_position: Optional[torch.LongTensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        **kwargs,
    ) -> Union[Tuple, Mamba2Output]:
        output_hidden_states = (
            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
        )
        use_cache = use_cache if use_cache is not None else (self.config.use_cache if not self.training else False)
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        if (input_ids is None) ^ (inputs_embeds is not None):  # ^ is python for xor
            raise ValueError("You must specify exactly one of input_ids or inputs_embeds")

        if inputs_embeds is None:
            inputs_embeds = self.embeddings(input_ids)

        if self.gradient_checkpointing and self.training and use_cache:
            use_cache = False

        if use_cache:
            if cache_params is None:
                cache_params = Mamba2Cache(
                    self.config, inputs_embeds.size(0), device=inputs_embeds.device, dtype=inputs_embeds.dtype
                )
                cache_position = torch.arange(0, self.config.conv_kernel, device=inputs_embeds.device)
            elif cache_position is None:
                # cases when we do manual forward instead of using `model.generate` which will initiate
                # `cache_position` and makes sure it is not None, throw error here instead of doing some
                # hack to conjecture the current cache position
                raise ValueError(
                    "You have to specify the `cache_position` manually when `use_cache=True` and `cache_params` is passed, "
                    "you don't have to pass a `cache_params` if you are in prefilling stage because in that case it will "
                    "be initialized for you automatically"
                )
        else:
            cache_params = None

        hidden_states = inputs_embeds
        all_hidden_states = () if output_hidden_states else None
        for mixer_block in self.layers:
            if self.gradient_checkpointing and self.training:
                hidden_states = self._gradient_checkpointing_func(
                    mixer_block.__call__, hidden_states, cache_params, cache_position, attention_mask
                )
            else:
                hidden_states = mixer_block(
                    hidden_states,
                    cache_params=cache_params,
                    cache_position=cache_position,
                    attention_mask=attention_mask,
                )

            if output_hidden_states:
                all_hidden_states = all_hidden_states + (hidden_states,)

        hidden_states = self.norm_f(hidden_states)

        if output_hidden_states:
            all_hidden_states = all_hidden_states + (hidden_states,)

        if not return_dict:
            return tuple(v for v in [hidden_states, cache_params, all_hidden_states] if v is not None)

        return Mamba2Output(
            last_hidden_state=hidden_states,
            cache_params=cache_params if use_cache else None,
            hidden_states=all_hidden_states,
        )

class IBS2ForClassification(Mamba2PreTrainedModel):
    _tied_weights_keys = []

    def __init__(self, config):
        super().__init__(config)
        self.backbone = IBS2Model(config)
        self.cls_head = nn.Linear(config.hidden_size, config.num_classes, bias=False)
        # Initialize weights and apply final processing
        self.post_init()
    
    def forward(
        self,
        input_ids: Optional[torch.LongTensor] = None,
        inputs_embeds: Optional[torch.FloatTensor] = None,
        cache_params: Optional[Mamba2Cache] = None,
        labels: Optional[torch.LongTensor] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        use_cache: Optional[bool] = None,
        cache_position: Optional[torch.Tensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        **kwargs,  # for now we need this for generation
    ):
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        mamba2_outputs = self.backbone(
            input_ids,
            cache_params=cache_params,
            inputs_embeds=inputs_embeds,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            use_cache=use_cache,
            cache_position=cache_position,
            attention_mask=attention_mask,
        )
        hidden_states = mamba2_outputs[0]

        logits = self.cls_head(hidden_states.to(self.cls_head.weight.dtype)).float()

        loss = None
        if labels is not None:
            labels = labels.to(logits.device)
            loss_fct = CrossEntropyLoss()
            loss = loss_fct(logits.view(-1, logits.size(-1)), labels.view(-1))

        if not return_dict:
            output = (logits,) + mamba2_outputs[1:]
            return ((loss,) + output) if loss is not None else output

        return loss
        # return Mamba2CausalLMOutput(
        #     loss=loss,
        #     logits=logits,
        #     cache_params=mamba2_outputs.cache_params,
        #     hidden_states=mamba2_outputs.hidden_states,
        # )

@add_start_docstrings(
    """
    The MAMBA2 Model transformer with a language modeling head on top (linear layer with weights not tied to the input
    embeddings).
    """,
    MAMBA2_START_DOCSTRING,
)
class IBS2ForCausalLM(Mamba2PreTrainedModel, GenerationMixin):
    _tied_weights_keys = []

    def __init__(self, config):
        super().__init__(config)
        self.backbone = IBS2Model(config)
        self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
        # Initialize weights and apply final processing
        self.post_init()

    def get_output_embeddings(self):
        return self.lm_head

    def set_output_embeddings(self, new_embeddings):
        self.lm_head = new_embeddings

    def get_input_embeddings(self):
        return self.backbone.get_input_embeddings()

    def set_input_embeddings(self, new_embeddings):
        return self.backbone.set_input_embeddings(new_embeddings)

    def prepare_inputs_for_generation(
        self,
        input_ids,
        inputs_embeds=None,
        use_cache=None,
        cache_params: Optional[Mamba2Cache] = None,
        cache_position: Optional[torch.LongTensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        **kwargs,
    ):
        # Overwitten -- uses `cache_params` as opposed to `past_key_values`

        if use_cache:
            # `cache_position` should have been initialized in `generate`
            if cache_position is None:
                raise ValueError(
                    "`cache_position` should not be None as it should have been initialized in "
                    "`model.generate`, you are responsible for passing in a valid `cache_position` if "
                    "you are calling `prepare_inputs_for_generation` directly with `use_cache=True`"
                )
            if cache_position[0] > 0:
                input_ids = input_ids[:, -1][..., None]

                if attention_mask is not None:
                    attention_mask = None
            else:
                # we initialize the `cache_position` to full size of `conv_states` at prefill stage
                # considering padding will be applied when input length is shorter, and truncation
                # will be applied when it is longer, so it will be equivalent to always have it match
                # the length of `cache_params.conv_states`, which is `config.conv_kernel`
                cache_position = torch.arange(0, self.config.conv_kernel, device=input_ids.device)

        if inputs_embeds is not None and cache_params is None:
            model_inputs = {"inputs_embeds": inputs_embeds}
        else:
            model_inputs = {"input_ids": input_ids}

        model_inputs.update(
            {
                "attention_mask": attention_mask,
                "cache_params": cache_params,
                "use_cache": use_cache,
                "cache_position": cache_position,
            }
        )
        return model_inputs

    @add_start_docstrings_to_model_forward(MAMBA2_INPUTS_DOCSTRING)
    @add_code_sample_docstrings(
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=Mamba2CausalLMOutput,
        config_class=_CONFIG_FOR_DOC,
    )
    def forward(
        self,
        input_ids: Optional[torch.LongTensor] = None,
        inputs_embeds: Optional[torch.FloatTensor] = None,
        cache_params: Optional[Mamba2Cache] = None,
        labels: Optional[torch.LongTensor] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        use_cache: Optional[bool] = None,
        cache_position: Optional[torch.Tensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        **kwargs,  # for now we need this for generation
    ) -> Union[Tuple, Mamba2CausalLMOutput]:
        r"""
        labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for language modeling. Note that the labels **are shifted** inside the model, i.e. you can set
            `labels = input_ids` Indices are selected in `[-100, 0, ..., config.vocab_size]` All labels set to `-100`
            are ignored (masked), the loss is only computed for labels in `[0, ..., config.vocab_size]`
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        mamba2_outputs = self.backbone(
            input_ids,
            cache_params=cache_params,
            inputs_embeds=inputs_embeds,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            use_cache=use_cache,
            cache_position=cache_position,
            attention_mask=attention_mask,
        )
        hidden_states = mamba2_outputs[0]

        logits = self.lm_head(hidden_states.to(self.lm_head.weight.dtype)).float()

        loss = None
        if labels is not None:
            # move labels to correct device to enable model parallelism
            labels = labels.to(logits.device)
            # Shift so that tokens < n predict n
            shift_logits = logits[..., :-1, :].contiguous()
            shift_labels = labels[..., 1:].contiguous()
            # Flatten the tokens
            loss_fct = CrossEntropyLoss()
            loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))

        if not return_dict:
            output = (logits,) + mamba2_outputs[1:]
            return ((loss,) + output) if loss is not None else output

        return Mamba2CausalLMOutput(
            loss=loss,
            logits=logits,
            cache_params=mamba2_outputs.cache_params,
            hidden_states=mamba2_outputs.hidden_states,
        )


__all__ = ["IBS2ForCausalLM", "IBS2Model", "Mamba2PreTrainedModel", "IBS2Block", "IBS2ForClassification"]