# Copyright 2023 The EASYDEL Author @erfanzar (Erfan Zare Chavoshi).
#
# 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
#
# https://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.
# coding=utf-8
# Copyright 2021 The Google Flax Team Authors and The 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.
import typing as tp
import chex
import jax
import jax.numpy as jnp
from flax import nnx as nn
from jax import lax
from easydel.infra.base_module import EasyDeLBaseModule
from easydel.infra.factory import TaskType, register_module
from easydel.infra.modeling_outputs import (
FlaxBaseModelOutputWithPastAndCrossAttentions,
FlaxCausalLMOutputWithCrossAttentions,
)
from easydel.infra.utils import (
ACT2FN,
auto_remat,
block_wise_ffn,
get_dot_general_by_bits,
)
from easydel.layers.attention import FlaxAttentionModule, FlexibleAttentionModule
from easydel.layers.caching import TransformerCache, TransformerCacheView
from easydel.modules.gpt2.gpt2_configuration import GPT2Config as GPT2Config
class Conv1D(nn.Module):
def __init__(
self,
in_features: int,
out_features: int,
use_bias: bool = True,
dtype: jnp.dtype = jnp.float32,
param_dtype: jnp.dtype = jnp.float32,
precision: jax.lax.PrecisionLike = None,
dot_general: tp.Optional[None] = None,
*,
rngs: nn.Rngs,
):
self.kernel = nn.Param(
nn.initializers.normal(stddev=0.02)(rngs.params(), (out_features, in_features)),
)
self.bias = nn.Param(
nn.initializers.zeros(
rngs.params(),
(in_features,),
)
if use_bias
else None
)
self.use_bias = use_bias
self.dtype = dtype
self.param_dtype = param_dtype
self.precision = precision
self.dot_general = dot_general
def __call__(self, inputs):
inputs = jnp.asarray(inputs, self.dtype)
bias = self.bias.value
kernel = self.kernel.value.transpose().astype(self.dtype)
if self.dot_general is not None:
dot_general = self.dot_general
else:
dot_general = lax.dot_general
y = dot_general(
inputs,
kernel,
(((inputs.ndim - 1,), (0,)), ((), ())),
precision=self.precision,
)
if bias is not None:
y = y + bias.astype(self.dtype)
return y
class GPT2Attention(FlaxAttentionModule):
def __init__(
self,
config: GPT2Config,
dtype: jnp.dtype = jnp.float32,
param_dtype: jnp.dtype = jnp.float32,
precision: jax.lax.PrecisionLike = None,
causal: bool = True,
is_cross_attention: bool = False,
*,
rngs: nn.Rngs,
):
super().__init__(config=config)
self.precision = precision
self.dtype = dtype
self.rngs = rngs
self.is_cross_attention = is_cross_attention
self.causal = causal
self.embed_dim = config.hidden_size
self.num_heads = config.num_attention_heads
self.head_dim = self.embed_dim // self.num_heads
if self.is_cross_attention:
self.c_attn = Conv1D(
self.embed_dim,
2 * self.embed_dim,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.q_attn = Conv1D(
self.embed_dim,
self.embed_dim,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
else:
self.c_attn = Conv1D(
self.embed_dim,
3 * self.embed_dim,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.c_proj = Conv1D(
self.embed_dim,
self.embed_dim,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.resid_dropout = nn.Dropout(rate=config.resid_pdrop, rngs=rngs)
self.attention_performer = FlexibleAttentionModule(
dropout_prob=config.attn_pdrop,
base_config=config,
softmax_scale=self.head_dim**-0.5,
)
def _split_heads(self, hidden_states):
return hidden_states.reshape(
hidden_states.shape[:2] + (self.num_heads, self.head_dim)
)
def _merge_heads(self, hidden_states):
"""
Merges the attention heads into a single hidden state tensor.
Args:
hidden_states (chex.Array): The hidden states with separate head dimensions.
Returns:
chex.Array: The hidden states with merged head dimensions.
"""
return hidden_states.reshape(hidden_states.shape[:2] + (self.embed_dim,))
def __call__(
self,
hidden_states: chex.Array,
key_value_states: chex.Array,
attention_mask: chex.Array,
causal_mask: tp.Optional[chex.Array] = None,
cache_view: tp.Optional[TransformerCacheView] = None,
output_attentions: bool = False,
):
is_cross_attention = key_value_states is not None
if not is_cross_attention:
qkv_out = self.c_attn(hidden_states)
query, key, value = jnp.split(qkv_out, 3, axis=2)
else:
q_out = self.q_attn(hidden_states)
(query,) = jnp.split(q_out, 1, axis=2)
kv_out = self.c_attn(key_value_states)
key, value = jnp.split(kv_out, 2, axis=2)
query = self._split_heads(query)
key = self._split_heads(key)
value = self._split_heads(value)
init_attention_bias = lambda: None # noqa
if self.causal:
(
key,
value,
attention_mask,
init_attention_bias,
) = self.concatenate(
query=query,
key=key,
cache_view=cache_view,
value=value,
attention_mask=attention_mask,
causal_mask=causal_mask,
fcm_mask=None,
)
attn = self.attention_performer.forward(
query_states=query,
key_states=key,
value_states=value,
init_bias=init_attention_bias,
attention_mask=attention_mask,
causal=self.causal,
dropout_rng=self.rngs.params(),
segment_ids=None,
)
attn_output = self.shard_attention_prod(self._merge_heads(attn.attention_outputs))
attn_output = self.c_proj(attn_output)
attn_output = self.resid_dropout(attn_output)
outputs = (
(attn_output, attn.attention_weights) if output_attentions else (attn_output,)
)
return outputs
class GPT2MLP(nn.Module):
def __init__(
self,
config: GPT2Config,
intermediate_size: int,
dtype: jnp.dtype = jnp.float32,
param_dtype: jnp.dtype = jnp.float32,
precision: tp.Optional[jax.lax.Precision] = None,
*,
rngs: nn.Rngs,
):
super().__init__()
self.config = config
self.precision = precision
self.dtype = dtype
self.rngs = rngs
embed_dim = config.hidden_size
self.c_fc = Conv1D(
embed_dim,
intermediate_size,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.c_proj = Conv1D(
intermediate_size,
embed_dim,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.act = ACT2FN[config.activation_function]
self.dropout = nn.Dropout(
rate=config.resid_pdrop,
rngs=rngs,
)
def __call__(self, hidden_states):
return self.dropout(self.c_proj(self.act(self.c_fc(hidden_states))))
class GPT2Block(nn.Module):
def __init__(
self,
config: GPT2Config,
dtype: jnp.dtype = jnp.float32,
param_dtype: jnp.dtype = jnp.float32,
precision: tp.Optional[jax.lax.Precision] = None,
*,
rngs: nn.Rngs,
):
super().__init__()
self.config = config
self.dtype = dtype
self.param_dtype = param_dtype
self.precision = precision
hidden_size = self.config.hidden_size
inner_dim = (
self.config.n_inner if self.config.n_inner is not None else 4 * hidden_size
)
self.ln_1 = nn.LayerNorm(
config.hidden_size,
epsilon=config.layer_norm_epsilon,
dtype=dtype,
param_dtype=param_dtype,
rngs=rngs,
)
attn_block = GPT2Attention
mlp_block = GPT2MLP
attn_block, mlp_block = auto_remat(
attn_block,
mlp_block,
policy=config.gradient_checkpointing,
)
self.attn = attn_block(
config,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.ln_2 = nn.LayerNorm(
config.hidden_size,
epsilon=config.layer_norm_epsilon,
dtype=dtype,
param_dtype=param_dtype,
rngs=rngs,
)
if config.add_cross_attention:
self.crossattention = attn_block(
config=config,
dtype=dtype,
causal=True,
is_cross_attention=True,
)
self.ln_cross_attn = nn.LayerNorm(
config.hidden_size,
epsilon=config.layer_norm_epsilon,
dtype=dtype,
param_dtype=param_dtype,
rngs=rngs,
)
self.mlp = mlp_block(
config=config,
intermediate_size=inner_dim,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
def __call__(
self,
hidden_states,
attention_mask=None,
causal_mask=None,
encoder_hidden_states: tp.Optional[chex.Array] = None,
encoder_attention_mask: tp.Optional[chex.Array] = None,
cache_view: tp.Optional[TransformerCacheView] = None,
output_attentions: bool = False,
):
residual = hidden_states
hidden_states = self.ln_1(hidden_states)
attn_outputs = self.attn(
hidden_states,
None,
attention_mask,
causal_mask,
cache_view,
output_attentions,
)
attn_output = attn_outputs[0]
outputs = attn_outputs[1:]
hidden_states = attn_output + residual
if encoder_hidden_states is not None:
if not hasattr(self, "crossattention"):
raise ValueError(
f"If `encoder_hidden_states` are passed, {self} has to be instantiated with "
"cross-attention layers by setting `config.add_cross_attention=True`"
)
residual = hidden_states
hidden_states = self.ln_cross_attn(hidden_states)
cross_attn_outputs = self.crossattention(
hidden_states,
encoder_hidden_states,
encoder_attention_mask,
causal_mask,
None,
output_attentions,
)
attn_output = cross_attn_outputs[0]
hidden_states = residual + attn_output
outputs = outputs + cross_attn_outputs[1:]
residual = hidden_states
hidden_states = self.ln_2(hidden_states)
if self.config.use_scan_mlp:
feed_forward_hidden_states = block_wise_ffn(
self.mlp,
hidden_states,
self.config.scan_mlp_chunk_size,
)
else:
feed_forward_hidden_states = self.mlp(hidden_states)
hidden_states = residual + feed_forward_hidden_states
outputs = (hidden_states,) + outputs
return outputs
[docs]@register_module(
TaskType.BASE_MODULE,
config=GPT2Config,
model_type="gpt2",
)
class GPT2Model(EasyDeLBaseModule):
def __init__(
self,
config: GPT2Config,
dtype: jnp.dtype = jnp.float32,
param_dtype: jnp.dtype = jnp.float32,
precision: jax.lax.PrecisionLike = None,
*,
rngs: nn.Rngs,
):
super().__init__(
config=config,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.embed_dim = self.config.hidden_size
self.wte = nn.Embed(
self.config.vocab_size,
self.embed_dim,
embedding_init=jax.nn.initializers.normal(stddev=self.config.initializer_range),
dtype=self.dtype,
rngs=rngs,
param_dtype=param_dtype,
)
self.wpe = nn.Embed(
self.config.max_position_embeddings,
self.embed_dim,
embedding_init=jax.nn.initializers.normal(stddev=self.config.initializer_range),
dtype=self.dtype,
param_dtype=param_dtype,
rngs=rngs,
)
self.dropout = nn.Dropout(rate=self.config.embd_pdrop, rngs=rngs)
self.h = [
GPT2Block(
self.config,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
for i in range(self.config.num_hidden_layers)
]
self.ln_f = nn.LayerNorm(
self.config.hidden_size,
epsilon=self.config.layer_norm_epsilon,
dtype=self.dtype,
param_dtype=param_dtype,
rngs=rngs,
)
def __call__(
self,
input_ids,
attention_mask: tp.Optional[chex.Array] = None,
position_ids: tp.Optional[chex.Array] = None,
encoder_hidden_states: tp.Optional[chex.Array] = None,
encoder_attention_mask: tp.Optional[chex.Array] = None,
past_key_values: tp.Optional[TransformerCache] = None,
output_attentions: bool = False,
output_hidden_states: bool = False,
return_dict: bool = True,
):
batch_size, sequence_length = input_ids.shape
if attention_mask is None:
attention_mask = jnp.ones((batch_size, sequence_length), "b1")
else:
if attention_mask.dtype != jnp.bool:
attention_mask = jnp.astype(attention_mask == 1, "b1")
if position_ids is None:
position_ids = jnp.broadcast_to(
jnp.clip(jnp.cumsum(attention_mask, axis=-1) - 1, a_min=0),
(batch_size, sequence_length),
).astype(jnp.int32)
inputs_embeds = self.wte(input_ids.astype("i4"))
position_embeds = self.wpe(position_ids.astype("i4"))
hidden_states = inputs_embeds + position_embeds
hidden_states = self.dropout(hidden_states)
all_attentions = () if output_attentions else None
all_hidden_states = () if output_hidden_states else None
all_cross_attentions = (
() if (output_attentions and encoder_hidden_states is not None) else None
)
if past_key_values is None:
past_key_values = TransformerCache.init_empty(len(self.h))
for idx, block in enumerate(self.h):
if output_hidden_states:
all_hidden_states += (hidden_states,)
layer_outputs = block(
hidden_states=hidden_states,
attention_mask=attention_mask,
causal_mask=self.causal_mask,
encoder_hidden_states=encoder_hidden_states,
encoder_attention_mask=encoder_attention_mask,
cache_view=past_key_values.views[idx],
output_attentions=output_attentions,
)
hidden_states = layer_outputs[0]
if output_attentions:
all_attentions += (layer_outputs[1],)
if encoder_hidden_states is not None:
all_cross_attentions += (layer_outputs[2],)
outputs = (
hidden_states,
all_hidden_states,
all_attentions,
all_cross_attentions,
)
hidden_states = outputs[0]
hidden_states = self.ln_f(hidden_states)
if output_hidden_states:
all_hidden_states += (hidden_states,)
outputs = (hidden_states, all_hidden_states, all_attentions, all_cross_attentions)
if not return_dict:
return tuple(v for v in outputs if v is not None)
return FlaxBaseModelOutputWithPastAndCrossAttentions(
last_hidden_state=hidden_states,
hidden_states=outputs[1],
attentions=outputs[2],
cross_attentions=outputs[3],
)
[docs]@register_module(
TaskType.CAUSAL_LM,
config=GPT2Config,
model_type="gpt2",
)
class GPT2LMHeadModel(EasyDeLBaseModule):
def __init__(
self,
config: GPT2Config,
dtype: jnp.dtype = jnp.float32,
param_dtype: jnp.dtype = jnp.float32,
precision: jax.lax.PrecisionLike = None,
*,
rngs: nn.Rngs,
):
super().__init__(
config=config,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.transformer = GPT2Model(
config=config,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
)
self.lm_head = nn.Linear(
config.hidden_size,
config.vocab_size,
use_bias=False,
dtype=dtype,
param_dtype=param_dtype,
precision=precision,
rngs=rngs,
kernel_init=jax.nn.initializers.normal(stddev=config.initializer_range),
**get_dot_general_by_bits(config.bits, config.easy_method),
)
def __call__(
self,
input_ids,
attention_mask: tp.Optional[chex.Array] = None,
position_ids: tp.Optional[chex.Array] = None,
encoder_hidden_states: tp.Optional[chex.Array] = None,
encoder_attention_mask: tp.Optional[chex.Array] = None,
past_key_values: tp.Optional[TransformerCache] = None,
output_attentions: bool = False,
output_hidden_states: bool = False,
return_dict: bool = True,
):
outputs = self.transformer(
input_ids=input_ids,
attention_mask=attention_mask,
position_ids=position_ids,
past_key_values=past_key_values,
encoder_hidden_states=encoder_hidden_states,
encoder_attention_mask=encoder_attention_mask,
output_attentions=output_attentions,
output_hidden_states=output_hidden_states,
return_dict=return_dict,
)
hidden_states = outputs[0]
if self.config.tie_word_embeddings:
lm_logits = jax.lax.dot_general(
hidden_states,
self.transformer.wte.embedding.value.T,
(((hidden_states.ndim - 1), (0,)), ((), ())),
)
else:
lm_logits = self.lm_head(hidden_states)
if not return_dict:
return (lm_logits,) + outputs[1:]
return FlaxCausalLMOutputWithCrossAttentions(
logits=lm_logits,
hidden_states=outputs.hidden_states,
attentions=outputs.attentions,
cross_attentions=outputs.cross_attentions,
)