Witllm/qwen/modeling_qwen.py

# Copyright (c) Alibaba Cloud.
#
# This source code is licensed under the license found in the
# LICENSE file in the root directory of this source tree.

import copy
import math
import inspect
from typing import TYPE_CHECKING, Optional, Tuple, Union, Callable, List, Any, Generator

import torch
import torch.nn.functional as F
import torch.utils.checkpoint

from torch.nn import CrossEntropyLoss
from transformers import PreTrainedTokenizer, GenerationConfig, StoppingCriteriaList
from transformers.generation.logits_process import LogitsProcessorList

if TYPE_CHECKING:
    from transformers.generation.streamers import BaseStreamer
from transformers.generation.utils import GenerateOutput
from transformers.modeling_outputs import (
    BaseModelOutputWithPast,
    CausalLMOutputWithPast,
)
from transformers.modeling_utils import PreTrainedModel
from transformers.utils import logging

from torch import nn
from einops import rearrange

from configuration_qwen import QWenConfig
from qwen_generation_utils import (
    HistoryType,
    make_context,
    decode_tokens,
    StopWordsLogitsProcessor,
)

import sys

sys.path.append("..")
from tools import show

logger = logging.get_logger(__name__)


class QWenAttention(nn.Module):
    def __init__(self, config, index):
        super().__init__()

        self.register_buffer("masked_bias", torch.tensor(-1e4), persistent=False)
        self.seq_length = config.seq_length

        self.hidden_size = config.hidden_size
        self.split_size = config.hidden_size
        self.num_heads = config.num_attention_heads
        self.head_dim = self.hidden_size // self.num_heads

        self.scale_attn_weights = True

        self.projection_size = config.kv_channels * config.num_attention_heads

        assert self.projection_size % config.num_attention_heads == 0
        self.hidden_size_per_attention_head = self.projection_size // config.num_attention_heads

        self.c_attn = nn.Linear(config.hidden_size, 3 * self.projection_size)
        self.c_proj = nn.Linear(config.hidden_size, self.projection_size, bias=not config.no_bias)

        self.use_dynamic_ntk = config.use_dynamic_ntk

        logn_list = [math.log(i, self.seq_length) if i > self.seq_length else 1 for i in range(1, 32768)]
        logn_tensor = torch.tensor(logn_list)[None, :, None, None]
        self.register_buffer("logn_tensor", logn_tensor, persistent=False)

        self.attn_dropout = nn.Dropout(config.attn_dropout_prob)
        self.softmax_in_fp32 = config.softmax_in_fp32 if hasattr(config, "softmax_in_fp32") else False
        self.use_cache_kernel = config.use_cache_kernel if hasattr(config, "use_cache_kernel") else False
        cache_dtype = torch.float
        self.cache_qmax = torch.tensor(torch.iinfo(torch.uint8).max, dtype=cache_dtype)
        self.cache_qmin = torch.tensor(torch.iinfo(torch.uint8).min, dtype=cache_dtype)
        self.index = index

    def _split_heads(self, tensor, num_heads, attn_head_size):
        new_shape = tensor.size()[:-1] + (num_heads, attn_head_size)
        tensor = tensor.view(new_shape)
        return tensor

    def _merge_heads(self, tensor, num_heads, attn_head_size):
        tensor = tensor.contiguous()
        new_shape = tensor.size()[:-2] + (num_heads * attn_head_size,)
        return tensor.view(new_shape)

    def forward(
        self,
        hidden_states: Optional[Tuple[torch.FloatTensor]],
        rotary_pos_emb_list: Optional[List[List[torch.Tensor]]] = None,
        layer_past: Optional[Tuple[torch.Tensor]] = None,
        attention_mask: Optional[torch.FloatTensor] = None,
    ):
        mixed_x_layer = self.c_attn(hidden_states)
        query, key, value = mixed_x_layer.split(self.split_size, dim=2)
        query = self._split_heads(query, self.num_heads, self.head_dim)
        key = self._split_heads(key, self.num_heads, self.head_dim)
        value = self._split_heads(value, self.num_heads, self.head_dim)

        rotary_pos_emb = rotary_pos_emb_list[0]
        rotary_pos_emb = [i[:, -query.shape[1] :, :, :] for i in rotary_pos_emb]
        rotary_pos_emb = (rotary_pos_emb,) * 2
        q_pos_emb, k_pos_emb = rotary_pos_emb
        # Slice the pos emb for current inference
        query = apply_rotary_pos_emb(query, q_pos_emb)
        key = apply_rotary_pos_emb(key, k_pos_emb)

        if layer_past is not None:
            past_key, past_value = layer_past[0], layer_past[1]

            key = torch.cat((past_key, key), dim=1)
            value = torch.cat((past_value, value), dim=1)

        present = (key, value)

        key_size = key.size(1)
        if key_size > self.seq_length and not self.training:
            seq_start = key.size(1) - query.size(1)
            seq_end = key.size(1)
            logn_tensor = self.logn_tensor[:, seq_start:seq_end, :, :].type_as(query)
            query = query * logn_tensor.expand_as(query)

        key_size = key.size(1)
        if query.size(1) == key_size:
            causal_mask = torch.tril(torch.ones((key_size, key_size), dtype=torch.bool, device=query.device)).view(
                1, 1, key_size, key_size
            )
        else:
            causal_mask = None
        query = query.permute(0, 2, 1, 3)
        key = key.permute(0, 2, 1, 3)
        value = value.permute(0, 2, 1, 3)

        if attention_mask is not None:
            attention_mask = attention_mask.expand(-1, -1, causal_mask.size(2), -1)
            if causal_mask is not None:
                attention_mask = attention_mask.masked_fill(~causal_mask, torch.finfo(query.dtype).min)
        else:
            attention_mask = causal_mask

        # qk = query @ key.transpose(-2, -1)
        # qk = qk[0]
        # show.DumpTensorToImage(qk,"q_matmul_k_layer_"+str(self.index)+".png")

        attn_output = F.scaled_dot_product_attention(query, key, value, attn_mask=attention_mask).transpose(1, 2)
        context_layer = self._merge_heads(attn_output, self.num_heads, self.head_dim)
        attn_output = self.c_proj(context_layer)
        outputs = (attn_output, present)
        return outputs


class QWenMLP(nn.Module):
    def __init__(self, config):
        super().__init__()
        ff_dim_in = config.intermediate_size // 2
        self.w1 = nn.Linear(config.hidden_size, ff_dim_in, bias=not config.no_bias)
        self.w2 = nn.Linear(config.hidden_size, ff_dim_in, bias=not config.no_bias)
        self.c_proj = nn.Linear(ff_dim_in, config.hidden_size, bias=not config.no_bias)

    def forward(self, hidden_states):
        a1 = self.w1(hidden_states)
        a2 = self.w2(hidden_states)
        intermediate_parallel = a1 * F.silu(a2)
        output = self.c_proj(intermediate_parallel)
        return output


class QWenBlock(nn.Module):
    def __init__(self, config, index):
        super().__init__()
        hidden_size = config.hidden_size

        self.ln_1 = RMSNorm(
            hidden_size,
            eps=config.layer_norm_epsilon,
        )
        self.attn = QWenAttention(config, index)
        self.ln_2 = RMSNorm(
            hidden_size,
            eps=config.layer_norm_epsilon,
        )
        self.mlp = QWenMLP(config)
        self.index = index

    def forward(
        self,
        hidden_states: Optional[Tuple[torch.FloatTensor]],
        rotary_pos_emb_list: Optional[List[List[torch.Tensor]]] = None,
        layer_past: Optional[Tuple[torch.Tensor]] = None,
        attention_mask: Optional[torch.FloatTensor] = None,
    ):
        layernorm_output = self.ln_1(hidden_states)

        attn_outputs = self.attn(
            layernorm_output,
            rotary_pos_emb_list,
            layer_past=layer_past,
            attention_mask=attention_mask,
        )
        attn_output = attn_outputs[0]
        outputs = attn_outputs[1:]
        residual = hidden_states
        layernorm_input = attn_output + residual

        layernorm_output = self.ln_2(layernorm_input)
        residual = layernorm_input
        mlp_output = self.mlp(layernorm_output)
        hidden_states = residual + mlp_output
        outputs = (hidden_states,) + outputs
        return outputs


class QWenPreTrainedModel(PreTrainedModel):
    config_class = QWenConfig
    base_model_prefix = "transformer"
    is_parallelizable = False
    supports_gradient_checkpointing = True
    _no_split_modules = ["QWenBlock"]
    _skip_keys_device_placement = "past_key_values"

    def __init__(self, *inputs, **kwargs):
        super().__init__(*inputs, **kwargs)


class QWenModel(QWenPreTrainedModel):
    def __init__(self, config):
        super().__init__(config)
        self.vocab_size = config.vocab_size
        self.num_hidden_layers = config.num_hidden_layers
        self.embed_dim = config.hidden_size

        self.use_dynamic_ntk = config.use_dynamic_ntk
        self.seq_length = config.seq_length

        self.wte = nn.Embedding(self.vocab_size, self.embed_dim)

        self.drop = nn.Dropout(config.emb_dropout_prob)

        if config.rotary_pct == 1.0:
            self.rotary_ndims = None
        else:
            assert config.rotary_pct < 1
            self.rotary_ndims = int(config.kv_channels * config.rotary_pct)
        dim = self.rotary_ndims if self.rotary_ndims is not None else config.kv_channels
        self.rotary_emb = RotaryEmbedding(dim, base=config.rotary_emb_base)

        self.h = nn.ModuleList([QWenBlock(config, i) for i in range(config.num_hidden_layers)])
        self.ln_f = RMSNorm(
            self.embed_dim,
            eps=config.layer_norm_epsilon,
        )

        self.post_init()

    def get_ntk_alpha(self, true_seq_len):
        context_value = math.log(true_seq_len / self.seq_length, 2) + 1
        ntk_alpha = 2 ** math.ceil(context_value) - 1
        ntk_alpha = max(ntk_alpha, 1)
        return ntk_alpha

    def forward(
        self,
        input_ids: Optional[torch.LongTensor] = None,
        past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
        attention_mask: Optional[torch.FloatTensor] = None,
        head_mask: Optional[torch.FloatTensor] = None,
        inputs_embeds: Optional[torch.FloatTensor] = None,
    ):
        if input_ids is not None and inputs_embeds is not None:
            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
        elif input_ids is not None:
            input_shape = input_ids.size()
            input_ids = input_ids.view(-1, input_shape[-1])
            batch_size = input_ids.shape[0]
        elif inputs_embeds is not None:
            input_shape = inputs_embeds.size()[:-1]
            batch_size = inputs_embeds.shape[0]
        else:
            raise ValueError("You have to specify either input_ids or inputs_embeds")

        if past_key_values is None:
            past_key_values = tuple([None] * len(self.h))

        if attention_mask is not None:
            attention_mask = attention_mask.view(batch_size, -1)
            attention_mask = attention_mask[:, None, None, :]
            attention_mask = attention_mask.to(dtype=self.dtype)
            attention_mask = (1.0 - attention_mask) * torch.finfo(self.dtype).min

        head_mask = self.get_head_mask(head_mask, self.config.num_hidden_layers)

        if inputs_embeds is None:
            inputs_embeds = self.wte(input_ids)
        hidden_states = inputs_embeds

        kv_seq_len = hidden_states.size()[1]
        if past_key_values[0] is not None:
            # past key values[0][0] shape: bs * seq_len * head_num * dim
            kv_seq_len += past_key_values[0][0].shape[1]

        if self.training or not self.use_dynamic_ntk:
            ntk_alpha_list = [1.0]
        elif kv_seq_len != hidden_states.size()[1]:
            ntk_alpha_list = self.rotary_emb._ntk_alpha_cached_list
        else:
            ntk_alpha_list = []
            if attention_mask is not None and kv_seq_len > self.seq_length:
                true_seq_lens = attention_mask.squeeze(1).squeeze(1).eq(0).sum(dim=-1, dtype=torch.int32)
                for i in range(hidden_states.size()[0]):
                    true_seq_len = true_seq_lens[i].item()
                    ntk_alpha = self.get_ntk_alpha(true_seq_len)
                    ntk_alpha_list.append(ntk_alpha)
            else:
                ntk_alpha = self.get_ntk_alpha(kv_seq_len)
                ntk_alpha_list.append(ntk_alpha)
        self.rotary_emb._ntk_alpha_cached_list = ntk_alpha_list
        rotary_pos_emb_list = [self.rotary_emb(kv_seq_len, ntk_alpha=ntk_alpha) for ntk_alpha in ntk_alpha_list]

        hidden_states = self.drop(hidden_states)
        output_shape = input_shape + (hidden_states.size(-1),)

        presents = ()
        all_hidden_states = None
        for i, (block, layer_past) in enumerate(zip(self.h, past_key_values)):
            outputs = block(
                hidden_states,
                layer_past=layer_past,
                rotary_pos_emb_list=rotary_pos_emb_list,
                attention_mask=attention_mask,
            )
            hidden_states = outputs[0]
            presents = presents + (outputs[1],)

        hidden_states = self.ln_f(hidden_states)
        hidden_states = hidden_states.view(output_shape)
        return BaseModelOutputWithPast(
            last_hidden_state=hidden_states, past_key_values=presents, hidden_states=all_hidden_states
        )


class QWenLMHeadModel(QWenPreTrainedModel):
    def __init__(self, config):
        super().__init__(config)

        self.transformer = QWenModel(config)
        self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
        self.post_init()

    def prepare_inputs_for_generation(self, input_ids, past_key_values=None, inputs_embeds=None, **kwargs):
        if past_key_values:
            input_ids = input_ids[:, -1].unsqueeze(-1)

        if input_ids.size(0) == 1:
            attention_mask = None
        else:
            attention_mask = kwargs.get("attention_mask", None)

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

        model_inputs.update(
            {
                "past_key_values": past_key_values,
                "attention_mask": attention_mask,
            }
        )
        return model_inputs

    def forward(
        self,
        input_ids: Optional[torch.LongTensor] = None,
        past_key_values: Optional[Tuple[Tuple[torch.Tensor]]] = None,
        attention_mask: Optional[torch.FloatTensor] = None,
        head_mask: Optional[torch.FloatTensor] = None,
        inputs_embeds: Optional[torch.FloatTensor] = None,
        labels: Optional[torch.LongTensor] = None,
    ) -> Union[Tuple, CausalLMOutputWithPast]:
        transformer_outputs = self.transformer(
            input_ids,
            past_key_values=past_key_values,
            attention_mask=attention_mask,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
        )
        hidden_states = transformer_outputs[0]

        lm_logits = self.lm_head(hidden_states)

        loss = None
        if labels is not None:
            labels = labels.to(lm_logits.device)
            shift_logits = lm_logits[..., :-1, :].contiguous()
            shift_labels = labels[..., 1:].contiguous()
            loss_fct = CrossEntropyLoss()
            loss = loss_fct(shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1))

        # shift_labels = torch.ones([1,19]).to(lm_logits.device).to(torch.int64)
        # shift_logits = lm_logits[..., :-1, :].contiguous()
        # loss_fct = CrossEntropyLoss()
        # loss = loss_fct(
        #     shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1)
        # )
        # loss.backward()

        return CausalLMOutputWithPast(
            loss=loss,
            logits=lm_logits,
            past_key_values=transformer_outputs.past_key_values,
            hidden_states=transformer_outputs.hidden_states,
            attentions=transformer_outputs.attentions,
        )

    def chat(
        self,
        tokenizer: PreTrainedTokenizer,
        query: str,
        query_assistant: str,
        history: Optional[HistoryType],
        system: str = "You are a helpful assistant.",
        **kwargs,
    ) -> Tuple[str, HistoryType]:
        generation_config = self.generation_config

        if history is None:
            history = []
        else:
            history = copy.deepcopy(history)

        stop_words_ids = []

        max_window_size = kwargs.get("max_window_size", None)
        if max_window_size is None:
            max_window_size = generation_config.max_window_size
        raw_text, context_tokens = make_context(
            tokenizer, query, query_assistant, history=history, system=system, max_window_size=max_window_size
        )

        stop_words_ids.extend([[tokenizer.im_end_id], [tokenizer.im_start_id]])
        input_ids = torch.tensor([context_tokens]).to(self.device)
        outputs = self.generate(
            input_ids,
            stop_words_ids=stop_words_ids,
            **kwargs,
        )
        decoded, response, end_reason = decode_tokens(
            outputs[0],
            tokenizer,
            raw_text_len=len(raw_text),
            context_length=len(context_tokens),
            errors="replace",
        )
        history.append((query, response))
        return response, history, decoded

    def generate(
        self,
        inputs: Optional[torch.Tensor] = None,
        stop_words_ids=[],
        prefix_allowed_tokens_fn: Optional[Callable[[int, torch.Tensor], List[int]]] = None,
        **kwargs,
    ) -> Union[GenerateOutput, torch.LongTensor]:
        generation_config = self.generation_config

        # 1. Handle `generation_config` and kwargs that might update it, and validate the `.generate()` call
        self._validate_model_class()

        generation_config = copy.deepcopy(generation_config)
        model_kwargs = generation_config.update(**kwargs)  # All unused kwargs must be model kwargs
        generation_config.validate()
        self._validate_model_kwargs(model_kwargs.copy())

        # 2. Set generation parameters if not already defined

        if generation_config.pad_token_id is None and generation_config.eos_token_id is not None:
            if model_kwargs.get("attention_mask", None) is None:
                logger.warning(
                    "The attention mask and the pad token id were not set. As a consequence, you may observe "
                    "unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results."
                )
            eos_token_id = generation_config.eos_token_id
            if isinstance(eos_token_id, list):
                eos_token_id = eos_token_id[0]
            logger.warning(f"Setting `pad_token_id` to `eos_token_id`:{eos_token_id} for open-end generation.")
            generation_config.pad_token_id = eos_token_id

        # 3. Define model inputs
        inputs_tensor, model_input_name, model_kwargs = self._prepare_model_inputs(
            inputs, generation_config.bos_token_id, model_kwargs
        )
        # 4. Define other model kwargs

        accepts_attention_mask = "attention_mask" in set(inspect.signature(self.forward).parameters.keys())
        requires_attention_mask = "encoder_outputs" not in model_kwargs

        if model_kwargs.get("attention_mask", None) is None and requires_attention_mask and accepts_attention_mask:
            model_kwargs["attention_mask"] = self._prepare_attention_mask_for_generation(
                inputs_tensor,
                generation_config.pad_token_id,
                generation_config.eos_token_id,
            )

        # 5. Prepare `input_ids` which will be used for auto-regressive generation
        input_ids = inputs_tensor if model_input_name == "input_ids" else model_kwargs.pop("input_ids")

        # 6. Prepare `max_length` depending on other stopping criteria.
        input_ids_length = input_ids.shape[-1]
        has_default_max_length = kwargs.get("max_length") is None and generation_config.max_length is not None
        generation_config.max_length = generation_config.max_new_tokens + input_ids_length
        self._validate_generated_length(generation_config, input_ids_length, has_default_max_length)

        stop_words_logits_processor = StopWordsLogitsProcessor(
            stop_words_ids=stop_words_ids,
            eos_token_id=generation_config.eos_token_id,
        )
        logits_processor = LogitsProcessorList([stop_words_logits_processor])
        logits_processor = self._get_logits_processor(
            generation_config=generation_config,
            input_ids_seq_length=input_ids_length,
            encoder_input_ids=inputs_tensor,
            prefix_allowed_tokens_fn=prefix_allowed_tokens_fn,
            logits_processor=logits_processor,
            model_kwargs=model_kwargs,
            negative_prompt_ids=None,
            negative_prompt_attention_mask=None,
        )

        # 12. expand input_ids with `num_return_sequences` additional sequences per batch
        input_ids, model_kwargs = self._expand_inputs_for_generation(
            input_ids=input_ids,
            expand_size=generation_config.num_return_sequences,
            is_encoder_decoder=False,
            **model_kwargs,
        )

        # 13. run sample

        pad_token_id = generation_config.pad_token_id
        eos_token_id_tensor = torch.tensor([generation_config.eos_token_id]).to(input_ids.device)

        # init values
        stopping_criteria = self._get_stopping_criteria(
            generation_config=generation_config, stopping_criteria=StoppingCriteriaList()
        )

        logits_warper = self._get_logits_warper(generation_config)

        # init attention / hidden states / scores tuples
        scores = None

        # keep track of which sequences are already finished
        unfinished_sequences = torch.ones(input_ids.shape[0], dtype=torch.long, device=input_ids.device)

        this_peer_finished = False
        # auto-regressive generation
        while True:
            # prepare model inputs
            model_inputs = self.prepare_inputs_for_generation(input_ids, **model_kwargs)

            # forward pass to get next token
            outputs = self(**model_inputs)

            next_token_scores = outputs.logits[:, -1, :]

            # pre-process distribution
            next_token_scores = logits_processor(input_ids, next_token_scores)
            next_token_scores = logits_warper(input_ids, next_token_scores)

            # sample
            probs = nn.functional.softmax(next_token_scores, dim=-1)
            next_tokens = torch.multinomial(probs, num_samples=1).squeeze(1)

            next_tokens = next_tokens * unfinished_sequences + pad_token_id * (1 - unfinished_sequences)

            # update generated ids, model inputs, and length for next step
            input_ids = torch.cat([input_ids, next_tokens[:, None]], dim=-1)
            model_kwargs = self._update_model_kwargs_for_generation(outputs, model_kwargs, is_encoder_decoder=False)

            unfinished_sequences = unfinished_sequences.mul(
                next_tokens.tile(eos_token_id_tensor.shape[0], 1).ne(eos_token_id_tensor.unsqueeze(1)).prod(dim=0)
            )

            # stop when each sentence is finished
            if unfinished_sequences.max() == 0:
                this_peer_finished = True

            # stop if we exceed the maximum length
            if stopping_criteria(input_ids, scores):
                this_peer_finished = True

            if this_peer_finished:
                break

        return input_ids


class RotaryEmbedding(torch.nn.Module):
    def __init__(self, dim, base=10000):
        super().__init__()
        self.dim = dim
        self.base = base
        inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float() / dim))
        self.register_buffer("inv_freq", inv_freq, persistent=False)
        self._rotary_pos_emb_cache = None
        self._seq_len_cached = 0
        self._ntk_alpha_cached = 1.0
        self._ntk_alpha_cached_list = [1.0]

    def update_rotary_pos_emb_cache(self, seqlen, ntk_alpha=1.0):
        if seqlen > self._seq_len_cached or ntk_alpha != self._ntk_alpha_cached:
            base = self.base * ntk_alpha ** (self.dim / (self.dim - 2))
            self.inv_freq = 1.0 / (
                base ** (torch.arange(0, self.dim, 2, device=self.inv_freq.device).float() / self.dim)
            )
            self._seq_len_cached = max(2 * seqlen, 16)
            self._ntk_alpha_cached = ntk_alpha
            seq = torch.arange(self._seq_len_cached, device=self.inv_freq.device)
            freqs = torch.outer(seq.type_as(self.inv_freq), self.inv_freq)

            emb = torch.cat((freqs, freqs), dim=-1)
            emb = rearrange(emb, "n d -> 1 n 1 d")

            cos, sin = emb.cos(), emb.sin()
            self._rotary_pos_emb_cache = [cos, sin]

    def forward(self, max_seq_len, ntk_alpha=1.0):
        self.update_rotary_pos_emb_cache(max_seq_len, ntk_alpha)
        cos, sin = self._rotary_pos_emb_cache
        return [cos[:, :max_seq_len], sin[:, :max_seq_len]]


def _rotate_half(x):
    x = rearrange(x, "... (j d) -> ... j d", j=2)
    x1, x2 = x.unbind(dim=-2)
    return torch.cat((-x2, x1), dim=-1)


def apply_rotary_pos_emb(t, freqs):
    rot_dim = freqs[0].shape[-1]
    cos, sin = freqs
    t_float = t.float()
    t_rot, t_pass = t_float[..., :rot_dim], t_float[..., rot_dim:]
    t_rot = (t_rot * cos) + (_rotate_half(t_rot) * sin)
    return torch.cat((t_rot, t_pass), dim=-1).type_as(t)


class RMSNorm(torch.nn.Module):
    def __init__(self, dim: int, eps: float = 1e-6):
        super().__init__()
        self.eps = eps
        self.weight = nn.Parameter(torch.ones(dim))

    def _norm(self, x):
        return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)

    def forward(self, x):
        output = self._norm(x.float()).type_as(x)
        return output * self.weight