File size: 7,112 Bytes
1e2779e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 |
# coding=utf-8
# Copyright (c) 2025, Qwerky AI, Inc. 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.
"""QwerkyLlamaMambaHybrid model configuration"""
from typing import List, Optional
from transformers.configuration_utils import PretrainedConfig
from transformers.utils import logging
logger = logging.get_logger(__name__)
class QwerkyLlamaMambaHybridConfig(PretrainedConfig):
r"""
This is the configuration class to store the configuration of a [`MambaInLlamaMambaModel`]. It consolidates
both the transformer config and mamba config into a single configuration file.
Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs.
Read the documentation from [`PretrainedConfig`] for more information.
Args:
vocab_size (`int`, *optional*, defaults to 32000):
Vocabulary size of the MambaInLlama model.
hidden_size (`int`, *optional*, defaults to 4096):
Dimension of the hidden representations.
intermediate_size (`int`, *optional*, defaults to 11008):
Dimension of the MLP representations.
num_hidden_layers (`int`, *optional*, defaults to 32):
Number of hidden layers in the model.
num_attention_heads (`int`, *optional*, defaults to 32):
Number of attention heads for each attention layer.
num_key_value_heads (`int`, *optional*, defaults to 32):
Number of key-value heads for grouped query attention.
hidden_act (`str`, *optional*, defaults to "silu"):
The non-linear activation function in the MLP layers.
max_position_embeddings (`int`, *optional*, defaults to 2048):
The maximum sequence length that this model might ever be used with.
initializer_range (`float`, *optional*, defaults to 0.02):
The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
rms_norm_eps (`float`, *optional*, defaults to 1e-6):
The epsilon used by the rms normalization layers.
use_cache (`bool`, *optional*, defaults to `True`):
Whether or not the model should return the last key/values attentions.
pad_token_id (`int`, *optional*, defaults to 0):
The id of the padding token.
bos_token_id (`int`, *optional*, defaults to 1):
The id of the "beginning-of-sequence" token.
eos_token_id (`int`, *optional*, defaults to 2):
The id of the "end-of-sequence" token.
tie_word_embeddings (`bool`, *optional*, defaults to `False`):
Whether the model's input and output word embeddings should be tied.
rope_theta (`float`, *optional*, defaults to 10000.0):
The base period of the RoPE embeddings.
rope_scaling (`dict`, *optional*):
Dictionary containing the scaling configuration for the RoPE embeddings.
attention_dropout (`float`, *optional*, defaults to 0.0):
The dropout ratio for the attention probabilities.
# Mamba-specific config
d_model (`int`, *optional*):
Model dimension for Mamba layers. If not provided, defaults to `hidden_size`.
d_inner (`int`, *optional*):
Inner dimension for Mamba layers. If not provided, defaults to `intermediate_size`.
d_xb (`int`, *optional*, defaults to 2560):
Dimension for Mamba xB projection.
ssm_cfg (`dict`, *optional*, defaults to `{}`):
State space model configuration dictionary.
attn_layers (`List[int]`, *optional*, defaults to `[]`):
List of layer indices that use attention instead of Mamba.
"""
model_type = "qwerky_llama_mamba_hybrid"
keys_to_ignore_at_inference = ["past_key_values"]
def __init__(
self,
vocab_size: int = 32000,
hidden_size: int = 4096,
intermediate_size: int = 11008,
num_hidden_layers: int = 32,
num_attention_heads: int = 32,
num_key_value_heads: Optional[int] = None,
hidden_act: str = "silu",
max_position_embeddings: int = 2048,
initializer_range: float = 0.02,
rms_norm_eps: float = 1e-6,
use_cache: bool = True,
pad_token_id: int = 0,
bos_token_id: int = 1,
eos_token_id: int = 2,
tie_word_embeddings: bool = False,
rope_theta: float = 10000.0,
rope_scaling: Optional[dict] = None,
attention_dropout: float = 0.0,
# Mamba-specific parameters
d_model: Optional[int] = None,
d_inner: Optional[int] = None,
d_xb: int = 2560,
ssm_cfg: Optional[dict] = None,
attn_layers: Optional[List[int]] = None,
**kwargs,
):
self.vocab_size = vocab_size
self.hidden_size = hidden_size
self.intermediate_size = intermediate_size
self.num_hidden_layers = num_hidden_layers
self.num_attention_heads = num_attention_heads
self.num_key_value_heads = (
num_key_value_heads
if num_key_value_heads is not None
else num_attention_heads
)
self.hidden_act = hidden_act
self.max_position_embeddings = max_position_embeddings
self.initializer_range = initializer_range
self.rms_norm_eps = rms_norm_eps
self.use_cache = use_cache
self.rope_theta = rope_theta
self.rope_scaling = rope_scaling
self.attention_dropout = attention_dropout
# Mamba-specific parameters
self.d_model = d_model if d_model is not None else hidden_size
self.d_inner = d_inner if d_inner is not None else intermediate_size
self.d_xb = d_xb
self.ssm_cfg = ssm_cfg if ssm_cfg is not None else {}
self.attn_layers = attn_layers if attn_layers is not None else []
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,
)
# Set auto_map for external code loading
if "auto_map" not in kwargs:
self.auto_map = {
"AutoConfig": "configuration_qwerky_llama_mamba_hybrid.QwerkyLlamaMambaHybridConfig",
"AutoModelForCausalLM": "modeling_qwerky_llama_mamba_hybrid.QwerkyLlamaMambaHybridForCausalLM",
}
# Set architectures field
if "architectures" not in kwargs:
self.architectures = ["QwerkyLlamaMambaHybridForCausalLM"]
|