# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨

# This file was automatically generated from src/transformers/models/phi3/modular_phi3.py.

# Do NOT edit this file manually as any edits will be overwritten by the generation of

# the file from the modular. If any change should be done, please apply the change to the

# modular_phi3.py file directly. One of our CI enforces this.

# 🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨

# Copyright 2024 Microsoft and the HuggingFace Inc. 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.

from collections.abc import Callable

import torch

from torch import nn

from ...activations import ACT2FN

from ...cache_utils import Cache, DynamicCache

from ...generation import GenerationMixin

from ...integrations import use_kernel_forward_from_hub

from ...masking_utils import create_causal_mask, create_sliding_window_causal_mask

from ...modeling_flash_attention_utils import FlashAttentionKwargs

from ...modeling_layers import (

GenericForSequenceClassification,

GenericForTokenClassification,

GradientCheckpointingLayer,

)

from ...modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast

from ...modeling_rope_utils import ROPE_INIT_FUNCTIONS, dynamic_rope_update

from ...modeling_utils import ALL_ATTENTION_FUNCTIONS, PreTrainedModel

from ...processing_utils import Unpack

from ...utils import TransformersKwargs, auto_docstring, can_return_tuple

from ...utils.deprecation import deprecate_kwarg

from ...utils.generic import maybe_autocast, merge_with_config_defaults

from ...utils.output_capturing import capture_outputs

from .configuration_phi3 import Phi3Config

class Phi3MLP(nn.Module):

def __init__(self, config):

super().__init__()

self.config = config

self.gate_up_proj = nn.Linear(config.hidden_size, 2 * config.intermediate_size, bias=False)

self.down_proj = nn.Linear(config.intermediate_size, config.hidden_size, bias=False)

self.activation_fn = ACT2FN[config.hidden_act]

def forward(self, hidden_states: torch.FloatTensor) -> torch.FloatTensor:

up_states = self.gate_up_proj(hidden_states)

gate, up_states = up_states.chunk(2, dim=-1)

up_states = up_states * self.activation_fn(gate)

return self.down_proj(up_states)

class Phi3RotaryEmbedding(nn.Module):

@deprecate_kwarg("device", version="5.18")

def __init__(self, config: Phi3Config, device=None):

super().__init__()

self.max_seq_len_cached = config.max_position_embeddings

self.original_max_seq_len = config.max_position_embeddings

self.config = config

self.rope_type = self.config.rope_parameters["rope_type"]

rope_init_fn: Callable = self.compute_default_rope_parameters

if self.rope_type != "default":

rope_init_fn = ROPE_INIT_FUNCTIONS[self.rope_type]

inv_freq, self.attention_scaling = rope_init_fn(self.config, device)

self.inv_freq = nn.Buffer(inv_freq, persistent=False)

self.original_inv_freq = nn.Buffer(inv_freq.clone(), persistent=False)

@staticmethod

@deprecate_kwarg("device", version="5.18")

def compute_default_rope_parameters(config: Phi3Config, device=None, **kwargs) -> tuple[torch.Tensor, float]:

"""

Computes the inverse frequencies according to the original RoPE implementation

Args:

config ([`~transformers.PreTrainedConfig`]):

The model configuration.

Returns:

Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the

post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE).

"""

base = config.rope_parameters["rope_theta"]

partial_rotary_factor = config.rope_parameters.get("partial_rotary_factor", 1.0)

head_dim = getattr(config, "head_dim", None) or config.hidden_size // config.num_attention_heads

dim = int(head_dim * partial_rotary_factor)

attention_factor = 1.0 # Unused in this type of RoPE

# Compute the inverse frequencies

inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2, dtype=torch.float) / dim))

return inv_freq.to(device), attention_factor

@torch.no_grad()

@dynamic_rope_update # power user: used with advanced RoPE types (e.g. dynamic rope)

def forward(self, x, position_ids):

inv_freq_expanded = (

self.inv_freq[None, :, None].expand(position_ids.shape[0], -1, 1).to(dtype=torch.float, device=x.device)

)

position_ids_expanded = position_ids[:, None, :].float()

device_type = x.device.type if isinstance(x.device.type, str) and x.device.type != "mps" else "cpu"

# Disable any outside autocast context if any, to really force fp32

with maybe_autocast(device_type=device_type, enabled=False):

freqs = (inv_freq_expanded @ position_ids_expanded).transpose(1, 2)

emb = torch.cat((freqs, freqs), dim=-1)

cos = emb.cos() * self.attention_scaling

sin = emb.sin() * self.attention_scaling

return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)

def rotate_half(x):

"""Rotates half the hidden dims of the input."""

x1 = x[..., : x.shape[-1] // 2]

x2 = x[..., x.shape[-1] // 2 :]

return torch.cat((-x2, x1), dim=-1)

def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor:

"""

This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,

num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)

"""

batch, num_key_value_heads, slen, head_dim = hidden_states.shape

if n_rep == 1:

return hidden_states

hidden_states = hidden_states[:, :, None, :, :].expand(batch, num_key_value_heads, n_rep, slen, head_dim)

return hidden_states.reshape(batch, num_key_value_heads * n_rep, slen, head_dim)

def eager_attention_forward(

module: nn.Module,

query: torch.Tensor,

key: torch.Tensor,

value: torch.Tensor,

attention_mask: torch.Tensor | None,

scaling: float,

dropout: float = 0.0,

**kwargs: Unpack[TransformersKwargs],

):

key_states = repeat_kv(key, module.num_key_value_groups)

value_states = repeat_kv(value, module.num_key_value_groups)

attn_weights = torch.matmul(query, key_states.transpose(2, 3)) * scaling

if attention_mask is not None:

attn_weights = attn_weights + attention_mask

attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(query.dtype)

attn_weights = nn.functional.dropout(attn_weights, p=dropout, training=module.training)

attn_output = torch.matmul(attn_weights, value_states)

attn_output = attn_output.transpose(1, 2).contiguous()

return attn_output, attn_weights

def apply_rotary_pos_emb(q, k, cos, sin, unsqueeze_dim=1):

"""Applies Rotary Position Embedding to the query and key tensors.

Args:

q (`torch.Tensor`): The query tensor.

k (`torch.Tensor`): The key tensor.

cos (`torch.Tensor`): The cosine part of the rotary embedding.

sin (`torch.Tensor`): The sine part of the rotary embedding.

unsqueeze_dim (`int`, *optional*, defaults to 1):

The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and

sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note

that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and

k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes

cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have

the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2.

Returns:

`tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding.

"""

cos = cos.unsqueeze(unsqueeze_dim)

sin = sin.unsqueeze(unsqueeze_dim)

rotary_dim = cos.shape[-1]

q_rot, q_pass = q[..., :rotary_dim], q[..., rotary_dim:]

k_rot, k_pass = k[..., :rotary_dim], k[..., rotary_dim:]

q_embed = torch.cat([(q_rot * cos) + (rotate_half(q_rot) * sin), q_pass], dim=-1)

k_embed = torch.cat([(k_rot * cos) + (rotate_half(k_rot) * sin), k_pass], dim=-1)

return q_embed, k_embed

class Phi3Attention(nn.Module):

"""Multi-headed attention from 'Attention Is All You Need' paper"""

def __init__(self, config: Phi3Config, layer_idx: int | None = None):

super().__init__()

self.config = config

self.layer_idx = layer_idx

self.head_dim = getattr(config, "head_dim", config.hidden_size // config.num_attention_heads)

self.num_key_value_groups = config.num_attention_heads // config.num_key_value_heads

self.num_key_value_heads = config.num_key_value_heads

self.scaling = self.head_dim**-0.5

self.attention_dropout = config.attention_dropout

self.is_causal = True

op_size = config.num_attention_heads * self.head_dim + 2 * (config.num_key_value_heads * self.head_dim)

self.o_proj = nn.Linear(config.num_attention_heads * self.head_dim, config.hidden_size, bias=False)

self.qkv_proj = nn.Linear(config.hidden_size, op_size, bias=False)

def forward(

self,

hidden_states: torch.Tensor,

position_embeddings: tuple[torch.Tensor, torch.Tensor],

attention_mask: torch.Tensor | None,

past_key_values: Cache | None = None,

**kwargs: Unpack[FlashAttentionKwargs],

) -> tuple[torch.Tensor, torch.Tensor | None, tuple[torch.Tensor] | None]:

input_shape = hidden_states.shape[:-1]

hidden_shape = (*input_shape, -1, self.head_dim)

qkv = self.qkv_proj(hidden_states)

query_pos = self.config.num_attention_heads * self.head_dim

query_states = qkv[..., :query_pos]

key_states = qkv[..., query_pos : query_pos + self.num_key_value_heads * self.head_dim]

value_states = qkv[..., query_pos + self.num_key_value_heads * self.head_dim :]

query_states = query_states.view(hidden_shape).transpose(1, 2)

key_states = key_states.view(hidden_shape).transpose(1, 2)

value_states = value_states.view(hidden_shape).transpose(1, 2)

cos, sin = position_embeddings

query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)

if past_key_values is not None:

key_states, value_states = past_key_values.update(key_states, value_states, self.layer_idx)

attention_interface: Callable = ALL_ATTENTION_FUNCTIONS.get_interface(

self.config._attn_implementation, eager_attention_forward

)

attn_output, attn_weights = attention_interface(

self,

query_states,

key_states,

value_states,

attention_mask,

dropout=0.0 if not self.training else self.attention_dropout,

scaling=self.scaling,

sliding_window=getattr(self.config, "sliding_window", None),

**kwargs,

)

attn_output = attn_output.reshape(*input_shape, -1).contiguous()

attn_output = self.o_proj(attn_output)

return attn_output, attn_weights

@use_kernel_forward_from_hub("RMSNorm")

class Phi3RMSNorm(nn.Module):

def __init__(self, hidden_size, eps: float = 1e-6) -> None:

"""

Phi3RMSNorm is equivalent to T5LayerNorm

"""

super().__init__()

self.weight = nn.Parameter(torch.ones(hidden_size))

self.variance_epsilon = eps

def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:

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)

def extra_repr(self):

return f"{tuple(self.weight.shape)}, eps={self.variance_epsilon}"

class Phi3DecoderLayer(GradientCheckpointingLayer):

def __init__(self, config: Phi3Config, layer_idx: int):

super().__init__()

self.hidden_size = config.hidden_size

self.self_attn = Phi3Attention(config=config, layer_idx=layer_idx)

self.mlp = Phi3MLP(config)

self.input_layernorm = Phi3RMSNorm(config.hidden_size, eps=config.rms_norm_eps)

self.post_attention_layernorm = Phi3RMSNorm(config.hidden_size, eps=config.rms_norm_eps)

self.config = config

self.resid_attn_dropout = nn.Dropout(config.resid_pdrop)

self.resid_mlp_dropout = nn.Dropout(config.resid_pdrop)

def forward(

self,

hidden_states: torch.Tensor,

attention_mask: torch.Tensor | None = None,

position_ids: torch.LongTensor | None = None,

past_key_values: Cache | None = None,

use_cache: bool | None = False,

position_embeddings: tuple[torch.Tensor, torch.Tensor] | None = None,

**kwargs: Unpack[FlashAttentionKwargs],

) -> tuple[torch.FloatTensor, tuple[torch.FloatTensor, torch.FloatTensor] | None]:

residual = hidden_states

hidden_states = self.input_layernorm(hidden_states)

hidden_states, self_attn_weights = self.self_attn(

hidden_states=hidden_states,

attention_mask=attention_mask,

position_ids=position_ids,

past_key_values=past_key_values,

use_cache=use_cache,

position_embeddings=position_embeddings,

**kwargs,

)

hidden_states = residual + self.resid_attn_dropout(hidden_states) # main diff with Llama

residual = hidden_states

hidden_states = self.post_attention_layernorm(hidden_states)

hidden_states = self.mlp(hidden_states)

hidden_states = residual + self.resid_mlp_dropout(hidden_states) # main diff with Llama

return hidden_states

@auto_docstring

class Phi3PreTrainedModel(PreTrainedModel):

config: Phi3Config

base_model_prefix = "model"

supports_gradient_checkpointing = True

_no_split_modules = ["Phi3DecoderLayer"]

_skip_keys_device_placement = ["past_key_values"]

_supports_flash_attn = True

_supports_sdpa = True

_supports_flex_attn = True

_can_compile_fullgraph = True

_supports_attention_backend = True

_can_record_outputs = {

"hidden_states": Phi3DecoderLayer,

"attentions": Phi3Attention,

}

_version = "0.0.5"

@auto_docstring

class Phi3Model(Phi3PreTrainedModel):

def __init__(self, config: Phi3Config):

super().__init__(config)

self.padding_idx = config.pad_token_id

self.vocab_size = config.vocab_size

self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size, self.padding_idx)

self.layers = nn.ModuleList(

[Phi3DecoderLayer(config, layer_idx) for layer_idx in range(config.num_hidden_layers)]

)

self.norm = Phi3RMSNorm(config.hidden_size, eps=config.rms_norm_eps)

self.rotary_emb = Phi3RotaryEmbedding(config=config)

self.gradient_checkpointing = False

# Initialize weights and apply final processing

self.post_init()

@merge_with_config_defaults

@capture_outputs

@auto_docstring

def forward(

self,

input_ids: torch.LongTensor | None = None,

attention_mask: torch.Tensor | None = None,

position_ids: torch.LongTensor | None = None,

past_key_values: Cache | None = None,

inputs_embeds: torch.FloatTensor | None = None,

use_cache: bool | None = None,

**kwargs: Unpack[TransformersKwargs],

) -> BaseModelOutputWithPast:

if (input_ids is None) ^ (inputs_embeds is not None):

raise ValueError("You must specify exactly one of input_ids or inputs_embeds")

if inputs_embeds is None:

inputs_embeds = self.embed_tokens(input_ids)

if use_cache and past_key_values is None:

past_key_values = DynamicCache(config=self.config)

if position_ids is None:

past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0

position_ids = torch.arange(inputs_embeds.shape[1], device=inputs_embeds.device) + past_seen_tokens

position_ids = position_ids.unsqueeze(0)

mask_function = create_causal_mask if self.config.sliding_window is None else create_sliding_window_causal_mask

causal_mask = mask_function(

config=self.config,

inputs_embeds=inputs_embeds,

attention_mask=attention_mask,

past_key_values=past_key_values,

position_ids=position_ids,

)

hidden_states = inputs_embeds

position_embeddings = self.rotary_emb(hidden_states, position_ids=position_ids)

for decoder_layer in self.layers[: self.config.num_hidden_layers]:

hidden_states = decoder_layer(

hidden_states,

attention_mask=causal_mask,

position_ids=position_ids,

past_key_values=past_key_values,

use_cache=use_cache,

position_embeddings=position_embeddings,

**kwargs,

)

hidden_states = self.norm(hidden_states)

return BaseModelOutputWithPast(

last_hidden_state=hidden_states,

past_key_values=past_key_values if use_cache else None,

)

@auto_docstring

class Phi3ForCausalLM(Phi3PreTrainedModel, GenerationMixin):

_tied_weights_keys = {"lm_head.weight": "model.embed_tokens.weight"}

_tp_plan = {"lm_head": "colwise_gather_output"}

_pp_plan = {"lm_head": (["hidden_states"], ["logits"])}

_fsdp_plan = {"lm_head": "keep_full_weight"}

def __init__(self, config):

super().__init__(config)

self.model = Phi3Model(config)

self.vocab_size = config.vocab_size

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

# Initialize weights and apply final processing

self.post_init()

@can_return_tuple

@auto_docstring

def forward(

self,

input_ids: torch.LongTensor | None = None,

attention_mask: torch.Tensor | None = None,

position_ids: torch.LongTensor | None = None,

past_key_values: Cache | None = None,

inputs_embeds: torch.FloatTensor | None = None,

labels: torch.LongTensor | None = None,

use_cache: bool | None = None,

logits_to_keep: int | torch.Tensor = 0,

**kwargs: Unpack[TransformersKwargs],

) -> CausalLMOutputWithPast:

r"""

Example:

```python

>>> from transformers import AutoTokenizer, Phi3ForCausalLM

>>> model = Phi3ForCausalLM.from_pretrained("meta-phi3/Phi3-2-7b-hf")

>>> tokenizer = AutoTokenizer.from_pretrained("meta-phi3/Phi3-2-7b-hf")

>>> prompt = "Hey, are you conscious? Can you talk to me?"

>>> inputs = tokenizer(prompt, return_tensors="pt")

>>> # Generate

>>> generate_ids = model.generate(inputs.input_ids, max_length=30)

>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]

"Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."

```"""

outputs: BaseModelOutputWithPast = self.model(

input_ids=input_ids,

attention_mask=attention_mask,

position_ids=position_ids,

past_key_values=past_key_values,

inputs_embeds=inputs_embeds,

use_cache=use_cache,

**kwargs,

)

hidden_states = outputs.last_hidden_state

# Only compute necessary logits, and do not upcast them to float if we are not computing the loss

slice_indices = slice(-logits_to_keep, None) if isinstance(logits_to_keep, int) else logits_to_keep

logits = self.lm_head(hidden_states[:, slice_indices, :])

loss = None

if labels is not None:

loss = self.loss_function(logits=logits, labels=labels, vocab_size=self.config.vocab_size, **kwargs)

return CausalLMOutputWithPast(

loss=loss,

logits=logits,

past_key_values=outputs.past_key_values,

hidden_states=outputs.hidden_states,

attentions=outputs.attentions,

)

def prepare_inputs_for_generation(

self,

input_ids,

past_key_values=None,

attention_mask=None,

inputs_embeds=None,

position_ids=None,

use_cache=True,

logits_to_keep=0,

**kwargs,

):

# Overwritten -- this model may need to switch between short and long rope, invalidating the cache in the

# process

# When the first time input length reached long and short factor switching point, enforce re-compute cache

# It will cause downside of slower at this single token position, however, better than current failure.

if (

past_key_values

and hasattr(self.config, "original_max_position_embeddings")

and input_ids.shape[1] >= self.config.original_max_position_embeddings + 1

):

past_length = past_key_values.get_seq_length()

if past_length <= self.config.original_max_position_embeddings:

past_key_values = None

model_inputs = super().prepare_inputs_for_generation(

input_ids=input_ids,

past_key_values=past_key_values,

attention_mask=attention_mask,

inputs_embeds=inputs_embeds,

position_ids=position_ids,

use_cache=use_cache,

logits_to_keep=logits_to_keep,

**kwargs,

)

return model_inputs

class Phi3ForSequenceClassification(GenericForSequenceClassification, Phi3PreTrainedModel):

pass

class Phi3ForTokenClassification(GenericForTokenClassification, Phi3PreTrainedModel):

pass

__all__ = [

"Phi3PreTrainedModel",

"Phi3Model",

"Phi3ForCausalLM",

"Phi3ForSequenceClassification",

"Phi3ForTokenClassification",

]