text-generation-inference/server/text_generation_server/models/custom_modeling/phi_modeling.py

331 lines
11 KiB
Python
Raw Normal View History

# imlementation of the PhiModel and PhiForCausalLM classes
import torch
import torch.distributed
import math
from torch import nn
from typing import Optional, List, Tuple, Any
from transformers.configuration_utils import PretrainedConfig
from transformers.modeling_outputs import CausalLMOutputWithPast
Refactor layers. (#1866) # What does this PR do? <!-- Congratulations! You've made it this far! You're not quite done yet though. Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution. Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change. Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost. --> <!-- Remove if not applicable --> Fixes # (issue) ## Before submitting - [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case). - [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#start-contributing-pull-requests), Pull Request section? - [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link to it if that's the case. - [ ] Did you make sure to update the documentation with your changes? Here are the [documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and [here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation). - [ ] Did you write any new necessary tests? ## Who can review? Anyone in the community is free to review the PR once the tests have passed. Feel free to tag members/contributors who may be interested in your PR. <!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @ @OlivierDehaene OR @Narsil -->
2024-05-13 10:44:30 +00:00
from text_generation_server.layers import (
TensorParallelRowLinear,
TensorParallelColumnLinear,
TensorParallelEmbedding,
Revamp medusa implementation so that every model can benefit. (#1588) # What does this PR do? <!-- Congratulations! You've made it this far! You're not quite done yet though. Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution. Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change. Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost. --> <!-- Remove if not applicable --> Fixes # (issue) ## Before submitting - [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case). - [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#start-contributing-pull-requests), Pull Request section? - [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link to it if that's the case. - [ ] Did you make sure to update the documentation with your changes? Here are the [documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and [here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation). - [ ] Did you write any new necessary tests? ## Who can review? Anyone in the community is free to review the PR once the tests have passed. Feel free to tag members/contributors who may be interested in your PR. <!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @ @OlivierDehaene OR @Narsil -->
2024-02-26 18:49:28 +00:00
SpeculativeHead,
FastLinear,
)
# PhiConfig is the configuration class for the PhiModel.
class PhiConfig(PretrainedConfig):
def __init__(
self,
vocab_size=51200,
n_positions=2048,
n_embd=2560,
n_layer=32,
n_inner=None,
n_head=32,
rotary_dim=32,
layer_norm_epsilon=1e-5,
tie_word_embeddings=False,
pad_vocab_size_multiple=64,
pad_token_id=0,
bos_token_id=1,
eos_token_id=2,
no_bias=False,
**kwargs,
):
self.vocab_size = vocab_size
self.n_positions = n_positions
self.n_embd = n_embd
self.n_layer = n_layer
self.n_inner = n_inner
self.n_head = n_head
self.rotary_dim = rotary_dim
self.layer_norm_epsilon = layer_norm_epsilon
self.tie_word_embeddings = tie_word_embeddings
self.pad_vocab_size_multiple = pad_vocab_size_multiple
self.pad_token_id = pad_token_id
self.bos_token_id = bos_token_id
self.eos_token_id = eos_token_id
self.no_bias = no_bias
super().__init__(
pad_token_id=pad_token_id,
bos_token_id=bos_token_id,
eos_token_id=eos_token_id,
tie_word_embeddings=tie_word_embeddings,
**kwargs,
)
2024-01-26 18:04:57 +00:00
# RotaryEmbedding is a class that implements the rotary embedding.
class RotaryEmbedding(nn.Module):
def __init__(self, dim, max_seq_len):
super().__init__()
2024-01-26 18:04:57 +00:00
inv_freq = [1.0 / 10000.0 ** (i / dim) for i in range(0, dim, 2)]
inv_freq_len = len(inv_freq)
inv_freq = torch.tensor(inv_freq).view(1, inv_freq_len)
t = torch.arange(0, max_seq_len, dtype=torch.float).view(max_seq_len, 1)
freqs = t.matmul(inv_freq)
self.sin = freqs.sin()
self.cos = freqs.cos()
def apply_rotary_emb_qkv(self, qkv, seqlen_offset):
b_size, seqlen, three, _, _headdim = qkv.shape
if three != 3:
raise Exception("unexpected shape for qkv")
_, rotary_dim = self.cos.shape
rotary_dim = rotary_dim * 2
q_rot = qkv[:, :, 0, :, :rotary_dim]
q_pass = qkv[:, :, 0, :, rotary_dim:]
k_rot = qkv[:, :, 1, :, :rotary_dim]
k_pass = qkv[:, :, 1, :, rotary_dim:]
q12 = torch.chunk(q_rot, 2, dim=-1)
k12 = torch.chunk(k_rot, 2, dim=-1)
q1, q2 = q12[0], q12[1]
k1, k2 = k12[0], k12[1]
c = self.cos.narrow(0, seqlen_offset, seqlen).unsqueeze(1)
s = self.sin.narrow(0, seqlen_offset, seqlen).unsqueeze(1)
q_rot = torch.cat(
[
q1 * c - q2 * s,
q1 * s + q2 * c,
],
dim=-1,
)
k_rot = torch.cat(
[
k1 * c - k2 * s,
k1 * s + k2 * c,
],
dim=-1,
)
q = torch.cat([q_rot, q_pass], dim=-1)
k = torch.cat([k_rot, k_pass], dim=-1)
v = qkv[:, :, 2]
return q, k, v
# PhiCausalLMHead is the head of the PhiModel. It is a linear layer with a layer norm.
class PhiCausalLMHead(nn.Module):
def __init__(self, config, weights):
super().__init__()
self.ln = nn.LayerNorm.load(
prefix="lm_head.ln",
weights=weights,
eps=config.layer_norm_epsilon,
)
Revamp medusa implementation so that every model can benefit. (#1588) # What does this PR do? <!-- Congratulations! You've made it this far! You're not quite done yet though. Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution. Then, please replace this with a description of the change and which issue is fixed (if applicable). Please also include relevant motivation and context. List any dependencies (if any) that are required for this change. Once you're done, someone will review your PR shortly (see the section "Who can review?" below to tag some potential reviewers). They may suggest changes to make the code even better. If no one reviewed your PR after a week has passed, don't hesitate to post a new comment @-mentioning the same persons---sometimes notifications get lost. --> <!-- Remove if not applicable --> Fixes # (issue) ## Before submitting - [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case). - [ ] Did you read the [contributor guideline](https://github.com/huggingface/transformers/blob/main/CONTRIBUTING.md#start-contributing-pull-requests), Pull Request section? - [ ] Was this discussed/approved via a Github issue or the [forum](https://discuss.huggingface.co/)? Please add a link to it if that's the case. - [ ] Did you make sure to update the documentation with your changes? Here are the [documentation guidelines](https://github.com/huggingface/transformers/tree/main/docs), and [here are tips on formatting docstrings](https://github.com/huggingface/transformers/tree/main/docs#writing-source-documentation). - [ ] Did you write any new necessary tests? ## Who can review? Anyone in the community is free to review the PR once the tests have passed. Feel free to tag members/contributors who may be interested in your PR. <!-- Your PR will be replied to more quickly if you can figure out the right person to tag with @ @OlivierDehaene OR @Narsil -->
2024-02-26 18:49:28 +00:00
self.linear = SpeculativeHead.load(
config=config, prefix="lm_head.linear", weights=weights
)
def forward(self, hidden_states):
hidden_states = self.ln(hidden_states)
hidden_states = self.linear(hidden_states)
return hidden_states
2024-01-26 18:04:57 +00:00
# PhiMHA is a multi-head attention layer. This layer uses an attention mask to prevent tokens from attending to subsequent tokens.
class PhiMHA(nn.Module):
def __init__(self, prefix, config, weights):
super().__init__()
self.Wqkv = TensorParallelColumnLinear.load(
config, prefix=f"{prefix}.Wqkv", weights=weights, bias=not config.no_bias
)
self.out_proj = TensorParallelRowLinear.load(
config,
prefix=f"{prefix}.out_proj",
weights=weights,
bias=not config.no_bias,
)
self.op_size = config.n_embd
self.head_dim = int(config.n_embd / config.n_head)
self.num_heads = config.n_head
self.rotary_emb = RotaryEmbedding(
config.rotary_dim,
config.n_positions,
)
self.softmax_scale = 1.0 / math.sqrt(self.head_dim)
def forward(
self,
hidden_states,
past_kv_cache,
attention_mask=None,
):
b_size, seq_len, _n_embd = hidden_states.shape
qkv = self.Wqkv(hidden_states)
qkv = qkv.view(b_size, seq_len, 3, self.num_heads, self.head_dim)
seqlen_offset = 0 if past_kv_cache is None else past_kv_cache[0].shape[1]
q, k, v = self.rotary_emb.apply_rotary_emb_qkv(qkv, seqlen_offset)
# if there is a kv_cache, then we need to concatenate
if past_kv_cache is not None:
prev_k, prev_v = past_kv_cache
k = torch.cat([prev_k, k], dim=1)
v = torch.cat([prev_v, v], dim=1)
past_kv_cache = [k, v]
2024-01-26 18:04:57 +00:00
attn_weights = torch.einsum("bthd,bshd->bhts", q, k * self.softmax_scale)
if attention_mask is not None:
seqlen_k = k.shape[1]
seqlen_q = q.shape[1]
2024-01-26 18:04:57 +00:00
causal_mask = torch.triu(
torch.full((seqlen_q, seqlen_k), -10000.0, device=attn_weights.device),
1,
)
attn_weights = attn_weights + causal_mask.to(dtype=attn_weights.dtype)
2024-01-26 18:04:57 +00:00
attn_weights = torch.nn.functional.softmax(attn_weights, dim=-1)
attn_output = attn_weights.matmul(v.transpose(1, 2)).squeeze(0)
2024-01-26 18:04:57 +00:00
attn_output = (
attn_output.view((b_size, self.num_heads, seq_len, self.head_dim))
.transpose(1, 2)
.flatten(-2)
)
return self.out_proj(attn_output), past_kv_cache
2024-01-26 18:04:57 +00:00
# PhiMLP is a multi-layer perceptron. It contains two linear layers with a gelu activation function.
class PhiMLP(nn.Module):
def __init__(self, prefix, config, weights):
super().__init__()
self.n_inner = config.n_inner
self.fc1 = FastLinear.load(
config=config,
prefix=f"{prefix}.fc1",
weights=weights,
bias=False,
)
self.fc2 = FastLinear.load(
config=config,
prefix=f"{prefix}.fc2",
weights=weights,
bias=False,
)
self.activation = torch.nn.functional.gelu
2024-01-26 18:04:57 +00:00
def forward(self, hidden_states):
hidden_states = self.fc1(hidden_states)
hidden_states = self.activation(hidden_states)
hidden_states = self.fc2(hidden_states)
return hidden_states
2024-01-26 18:04:57 +00:00
# PhiBlock is a single transformer block. It contains a layer norm, a multi-head attention layer and an multi-layer perceptron.
class PhiBlock(nn.Module):
def __init__(self, layer_id, config, weights):
super().__init__()
self.layer_id = layer_id
2024-01-26 18:04:57 +00:00
self.layer_norm = nn.LayerNorm.load(
prefix=f"{layer_id}.ln", weights=weights, eps=config.layer_norm_epsilon
)
self.mixer = PhiMHA(prefix=f"{layer_id}.mixer", config=config, weights=weights)
self.mlp = PhiMLP(prefix=f"{layer_id}.mlp", config=config, weights=weights)
def forward(
self,
hidden_states,
kv_cache,
attention_mask,
):
residual = hidden_states
hidden_states = self.layer_norm(hidden_states)
2024-01-26 18:04:57 +00:00
attn_outputs, past_kv_cache = self.mixer(
hidden_states, kv_cache, attention_mask
)
feed_forward_hidden_states = self.mlp(hidden_states)
out = attn_outputs + feed_forward_hidden_states + residual
return out, past_kv_cache
2024-01-26 18:04:57 +00:00
# PhiModel implements the embedding layer and the transformer blocks.
class PhiModel(nn.Module):
def __init__(self, config, weights):
super().__init__()
self.tp_rank = weights.process_group.rank()
self.tp_world_size = weights.process_group.size()
self.embed_tokens = TensorParallelEmbedding(
prefix="transformer.embd.wte", weights=weights
2024-01-26 18:04:57 +00:00
)
self.blocks = nn.ModuleList(
2024-01-26 18:04:57 +00:00
[
PhiBlock(f"transformer.h.{layer_id}", config, weights)
for layer_id in range(config.n_layer)
]
)
def forward(
self,
input_ids: torch.LongTensor,
past_key_values: Optional[List[Tuple[torch.FloatTensor]]] = None,
attention_mask: Optional[torch.ByteTensor] = None,
return_dict: Optional[bool] = None,
use_cache: Optional[bool] = None,
) -> Tuple[torch.Tensor, List[Tuple[torch.Tensor, torch.Tensor]]]:
hidden_states = self.embed_tokens(input_ids)
seq_len = hidden_states.shape[1]
mask = None if seq_len <= 1 else attention_mask
2024-01-26 18:04:57 +00:00
past_key_values = (
[None] * len(self.blocks) if past_key_values is None else past_key_values
)
for index, block in enumerate(self.blocks):
2024-01-26 18:04:57 +00:00
hidden_states, new_key_values = block(
hidden_states, past_key_values[index], mask
)
past_key_values[index] = new_key_values
return hidden_states, past_key_values
2024-01-26 18:04:57 +00:00
# PhiForCausalLM wraps the PhiModel and PhiCausalLMHead together and returns a CausalLMOutputWithPast object.
class PhiForCausalLM(torch.nn.Module):
def __init__(self, config, weights):
super().__init__()
self.model = PhiModel(config, weights)
self.lm_head = PhiCausalLMHead(config, weights)
def forward(
self,
input_ids: torch.LongTensor,
past_key_values: Optional[List[Tuple[torch.FloatTensor]]] = None,
attention_mask: Optional[torch.ByteTensor] = None,
return_dict: Optional[bool] = None,
use_cache: Optional[bool] = None,
labels: Optional[torch.LongTensor] = None,
) -> Tuple[torch.Tensor, List[Tuple[torch.Tensor, torch.Tensor]]]:
model_output = self.model(
input_ids, past_key_values, attention_mask, return_dict, use_cache
)
logits = self.lm_head(model_output[0])
loss = None
if labels is not None:
loss = nn.CrossEntropyLoss()(
2024-01-26 18:04:57 +00:00
logits[:, :-1].view(-1, logits.size(-1)), labels[:, 1:].view(-1)
)
if not return_dict:
2024-01-26 18:04:57 +00:00
return (
((loss,) + (logits,) + model_output[1:])
if loss is not None
else (logits,) + model_output[1:]
)
return CausalLMOutputWithPast(
loss=loss,
logits=logits,
past_key_values=model_output[1],
hidden_states=None,
attentions=None,
)