|
|
|
|
|
from transformers.configuration_utils import PretrainedConfig |
|
from open_clip import get_model_config |
|
import math |
|
from typing import Optional |
|
|
|
class PhiConfig(PretrainedConfig): |
|
"""Phi configuration.""" |
|
|
|
model_type = "phi-msft" |
|
attribute_map = { |
|
"max_position_embeddings": "n_positions", |
|
"hidden_size": "n_embd", |
|
"num_attention_heads": "n_head", |
|
"num_hidden_layers": "n_layer", |
|
} |
|
|
|
def __init__( |
|
self, |
|
vocab_size: int = 51200, |
|
n_positions: int = 2048, |
|
n_embd: int = 1024, |
|
n_layer: int = 20, |
|
n_inner: Optional[int] = None, |
|
n_head: int = 16, |
|
n_head_kv: Optional[int] = None, |
|
rotary_dim: Optional[int] = 32, |
|
activation_function: Optional[str] = "gelu_new", |
|
flash_attn: bool = False, |
|
flash_rotary: bool = False, |
|
fused_dense: bool = False, |
|
attn_pdrop: float = 0.0, |
|
embd_pdrop: float = 0.0, |
|
resid_pdrop: float = 0.0, |
|
layer_norm_epsilon: float = 1e-5, |
|
initializer_range: float = 0.02, |
|
tie_word_embeddings: bool = False, |
|
pad_vocab_size_multiple: int = 64, |
|
**kwargs |
|
) -> None: |
|
self.vocab_size = int(math.ceil(vocab_size / pad_vocab_size_multiple) * pad_vocab_size_multiple) |
|
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.n_head_kv = n_head_kv |
|
self.rotary_dim = min(rotary_dim, n_embd // n_head) |
|
self.activation_function = activation_function |
|
self.flash_attn = flash_attn |
|
self.flash_rotary = flash_rotary |
|
self.fused_dense = fused_dense |
|
self.attn_pdrop = attn_pdrop |
|
self.embd_pdrop = embd_pdrop |
|
self.resid_pdrop = resid_pdrop |
|
self.layer_norm_epsilon = layer_norm_epsilon |
|
self.initializer_range = initializer_range |
|
|
|
super().__init__(tie_word_embeddings=tie_word_embeddings, **kwargs) |
|
|
|
|
|
class LlavaConfig(PretrainedConfig): |
|
model_type = "llava" |
|
is_composition = False |
|
|
|
def __init__( |
|
self, |
|
text_config=None, |
|
vision_tower_name="ViT-SO400M-14-SigLIP-384", |
|
ignore_index=-100, |
|
image_token_index=50297, |
|
projector_hidden_act="gelu", |
|
projector_tokens_num=1, |
|
vocab_size=51200, |
|
**kwargs, |
|
): |
|
self.ignore_index = ignore_index |
|
self.image_token_index = image_token_index |
|
self.projector_hidden_act = projector_hidden_act |
|
self.projector_tokens_num = projector_tokens_num |
|
self.vocab_size = vocab_size |
|
|
|
self.vision_tower_name = vision_tower_name |
|
vision_config = get_model_config(vision_tower_name) |
|
self.vision_embed_dim = vision_config["embed_dim"] |
|
|
|
self.vocab_size = self.vocab_size |
|
|
|
self.text_config = text_config |
|
if isinstance(self.text_config, dict): |
|
text_config["model_type"] = text_config["model_type"] if "model_type" in text_config else "llama" |
|
self.text_config = PhiConfig(**text_config) |
|
self.vocab_size = self.text_config.vocab_size |
|
|
|
super().__init__(**kwargs) |
