# Copyright 2025 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.
from eformer.common_types import ColumnWise, Replicated, RowWise
from easydel.infra.base_module import EasyDeLBaseConfig
from easydel.infra.etils import EasyDeLGradientCheckPointers
from easydel.infra.factory import register_config
from easydel.infra.utils import AttnMaskDetail, AttnMaskType
[docs]@register_config("cohere2")
class Cohere2Config(EasyDeLBaseConfig):
"""
Configuration objects inherit from [`EasyDeLBaseConfig`] and can be used to control the model outputs. Read
the documentation from [`EasyDeLBaseConfig`] for more information.
Args:
vocab_size (`int`, *optional*, defaults to 256000):
Vocabulary size of the Cohere model. Defines the number of different tokens that can be represented by the
`inputs_ids` passed to the forward method.
hidden_size (`int`, *optional*, defaults to 8192):
Dimensionality of the encoder layers and the pooler layer.
intermediate_size (`int`, *optional*, defaults to 22528):
Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
logit_scale (`float`, *optional*, defaults to 0.0625):
A logit scale value used in the attention layer.
num_hidden_layers (`int`, *optional*, defaults to 40):
Number of hidden layers in the Transformer encoder.
num_attention_heads (`int`, *optional*, defaults to 64):
Number of attention heads for each attention layer in the Transformer encoder.
num_key_value_heads (`int`, *optional*):
Number of key and value heads for each attention layer in the Transformer encoder. Will default to
`num_attention_heads` if not set.
hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
The non-linear activation function (function or string) to use in the encoder and pooler. If string,
`"gelu"`, `"relu"`, `"swish"` and `"gelu_new"` are supported.
max_position_embeddings (`int`, *optional*, defaults to 8192):
The maximum sequence length that this model might ever be used with. Typically set this to something large
just in case (e.g., 2048 or 4096).
initializer_range (`float`, *optional*, defaults to 0.02):
The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
layer_norm_eps (`float`, *optional*, defaults to 1e-5):
The epsilon used by the layer normalization layers.
use_cache (`bool`, *optional*, defaults to `True`):
Whether or not the model should return the last key/values attentions (not used by all models). Only
relevant if `config.is_decoder=True`.
pad_token_id (`int`, *optional*, defaults to 0):
The index of the padding token in the vocabulary.
bos_token_id (`int`, *optional*, defaults to 5):
The index of the beginning of sequence token in the vocabulary.
eos_token_id (`int`, *optional*, defaults to 255001):
The index of the end of sequence token in the vocabulary.
tie_word_embeddings (`bool`, *optional*, defaults to `True`):
Whether to tie the weights of the input embeddings and the output embeddings.
rope_theta (`float`, *optional*, defaults to 10000.0):
The theta value to use for rotary position embeddings.
attention_bias (`bool`, *optional*, defaults to `False`):
Whether to use attention bias.
attention_dropout (`float`, *optional*, defaults to 0.0):
The dropout ratio for the attention probabilities.
use_qk_norm (`bool`, *optional*, defaults to `False`):
Whether to use query and key normalization.
gradient_checkpointing (`str`, *optional*, defaults to `"nothing_saveable"`):
The gradient checkpointing configuration.
bits (`int`, *optional*):
The number of bits to quantize the model to.
"""
model_type: str = "cohere"
def __init__(
self,
vocab_size=256000,
hidden_size=8192,
intermediate_size=22528,
logit_scale=0.0625,
num_hidden_layers=40,
num_attention_heads=64,
num_key_value_heads=None,
hidden_act="silu",
max_position_embeddings=8192,
initializer_range=0.02,
layer_norm_eps=1e-5,
use_cache=True,
pad_token_id=0,
bos_token_id=5,
eos_token_id=255001,
tie_word_embeddings=True,
rope_theta=10000.0,
rope_scaling=None,
attention_bias=False,
attention_dropout=0.0,
sliding_window=4096,
sliding_window_pattern=4,
layer_types: list[str] | None = None,
gradient_checkpointing: EasyDeLGradientCheckPointers = EasyDeLGradientCheckPointers.NONE,
bits: int | None = None,
**kwargs,
):
"""Initializes the Cohere2Config instance.
Args:
vocab_size (int): Vocabulary size.
hidden_size (int): Dimensionality of the hidden layers.
intermediate_size (int): Dimensionality of the intermediate feed-forward layer.
logit_scale (float): Logit scale for attention.
num_hidden_layers (int): Number of hidden layers.
num_attention_heads (int): Number of attention heads.
num_key_value_heads (Optional[int]): Number of key/value heads.
hidden_act (str): Activation function.
max_position_embeddings (int): Maximum sequence length.
initializer_range (float): Initializer range for weights.
layer_norm_eps (float): Epsilon for layer normalization.
use_cache (bool): Whether to use caching.
pad_token_id (int): Padding token ID.
bos_token_id (int): Beginning of sequence token ID.
eos_token_id (int): End of sequence token ID.
tie_word_embeddings (bool): Whether to tie word embeddings.
rope_theta (float): RoPE theta value.
rope_scaling (Optional[dict]): RoPE scaling configuration.
attention_bias (bool): Whether to use attention bias.
attention_dropout (float): Dropout rate for attention.
sliding_window (int): Sliding window size for attention.
sliding_window_pattern (int): Pattern for sliding window attention (unused in current Flax implementation).
gradient_checkpointing (EasyDeLGradientCheckPointers): Gradient checkpointing strategy.
bits (Optional[int]): Number of bits for quantization.
**kwargs: Additional keyword arguments.
"""
self.vocab_size = vocab_size
self.max_position_embeddings = max_position_embeddings
self.hidden_size = hidden_size
self.logit_scale = logit_scale
self.intermediate_size = intermediate_size
self.num_hidden_layers = num_hidden_layers
self.num_attention_heads = num_attention_heads
if num_key_value_heads is None:
num_key_value_heads = num_attention_heads
self.num_key_value_heads = num_key_value_heads
self.hidden_act = hidden_act
self.initializer_range = initializer_range
self.layer_norm_eps = layer_norm_eps
self.use_cache = use_cache
self.rope_theta = rope_theta
self.rope_scaling = rope_scaling
self.attention_bias = attention_bias
self.attention_dropout = attention_dropout
self.sliding_window = sliding_window
self.sliding_window_pattern = sliding_window_pattern
self.head_dim = hidden_size // num_attention_heads
self.gradient_checkpointing = gradient_checkpointing
self.bits = bits
self.layer_types = layer_types
if self.layer_types is None:
self.layer_types = [
"sliding_attention" if bool((i + 1) % self.sliding_window_pattern) else "full_attention"
for i in range(self.num_hidden_layers)
]
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,
)
[docs] def get_partition_rules(self, *args, **kwargs):
"""
Get the partition rules for the model.
Returns:
`tp.Tuple[tp.Tuple[str, PartitionSpec]]`: The partition rules.
"""
pmag = self.partition_manager
return (
(r"embed_tokens/embedding", pmag.resolve(ColumnWise)),
(r"self_attn/(q_proj|k_proj|v_proj)/kernel", pmag.resolve(ColumnWise)),
(r"self_attn/o_proj/kernel", pmag.resolve(RowWise)),
(r"self_attn/.*proj/bias", pmag.resolve(Replicated)),
(r"mlp/(gate_proj|up_proj)/kernel", pmag.resolve(ColumnWise)),
(r"mlp/down_proj/kernel", pmag.resolve(RowWise)),
(r"mlp/.*proj/bias", pmag.resolve(Replicated)),
(r".*layernorm/kernel", pmag.resolve(Replicated)),
(r"norm/kernel", pmag.resolve(Replicated)),
(r"lm_head/kernel", pmag.resolve(ColumnWise)),
(r"score/kernel", pmag.resolve(RowWise)),
(r".*bias", pmag.resolve(Replicated)),
(r".*", pmag.resolve(Replicated)),
)
[docs] def get_mask_details(self) -> dict[int, AttnMaskDetail]:
"""Retrieve attention mask details for each layer in the model.
This method generates a dictionary mapping layer indices to their corresponding attention mask details.
If a sliding window is defined, each layer is assigned a sliding window attention mask with the specified size.
Returns:
dict[int, AttnMaskDetail]: A dictionary where keys are layer indices (int) and values are AttnMaskDetail
objects specifying the attention mask type and size for each layer.
Notes:
- If `self.sliding_window` is None, an empty dictionary is returned.
- The method iterates over `self.num_hidden_layers` to assign mask details for each layer.
- The attention mask type is set to `AttnMaskType.SLIDING` when a sliding window is defined.
"""
mapping = {}
if self.layer_types is not None:
for layer_idx in range(self.num_hidden_layers):
mapping[layer_idx] = AttnMaskDetail(
mask_type=AttnMaskType.from_hf(self.layer_types[layer_idx]),
size=self.sliding_window,
)
return mapping