File size: 7,758 Bytes
c4b099e |
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 |
"""TPU Gemma2 model configuration"""
from transformers.configuration_utils import PretrainedConfig
class TPUGemma2Config(PretrainedConfig):
r"""
This is the configuration class to store the configuration of a [`Gemma2Model`]. It is used to instantiate an Gemma2
model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
defaults will yield a similar configuration to that of the Gemma2-7B.
e.g. [google/gemma2-7b](https://huggingface.co/google/gemma2-7b)
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 256000):
Vocabulary size of the Gemma2 model. Defines the number of different tokens that can be represented by the
`inputs_ids` passed when calling [`Gemma2Model`]
hidden_size (`int`, *optional*, defaults to 3072):
Dimension of the hidden representations.
intermediate_size (`int`, *optional*, defaults to 24576):
Dimension of the MLP representations.
num_hidden_layers (`int`, *optional*, defaults to 28):
Number of hidden layers in the Transformer decoder.
num_attention_heads (`int`, *optional*, defaults to 16):
Number of attention heads for each attention layer in the Transformer decoder.
num_key_value_heads (`int`, *optional*, defaults to 16):
This is the number of key_value heads that should be used to implement Grouped Query Attention. If
`num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
`num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When
converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
by meanpooling all the original heads within that group. For more details checkout [this
paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
`num_attention_heads`.
head_dim (`int`, *optional*, defaults to 256):
The attention head dimension.
hidden_activation (`str` or `function`, *optional*, defaults to `"gelu_pytorch_tanh"`):
The non-linear activation function (function or string) in the decoder. Will default to `"gelu_pytorch_tanh"`
if not specified. `"gelu_pytorch_tanh"` uses an approximation of the `"gelu"` activation function.
max_position_embeddings (`int`, *optional*, defaults to 8192):
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-06):
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 (not used by all models). Only
relevant if `config.is_decoder=True`.
pad_token_id (`int`, *optional*, defaults to 0):
Padding token id.
eos_token_id (`int`, *optional*, defaults to 1):
End of stream token id.
bos_token_id (`int`, *optional*, defaults to 2):
Beginning of stream token id.
tie_word_embeddings (`bool`, *optional*, defaults to `True`):
Whether to tie weight embeddings
rope_theta (`float`, *optional*, defaults to 10000.0):
The base period of the RoPE embeddings.
attention_bias (`bool`, defaults to `False`, *optional*, defaults to `False`):
Whether to use a bias in the query, key, value and output projection layers during self-attention.
attention_dropout (`float`, *optional*, defaults to 0.0):
The dropout ratio for the attention probabilities.
query_pre_attn_scalar (`float`, *optional*, defaults to 224): scaling factor used on the attention scores
sliding_window (`int`, *optional*, defaults to 4096): in Gemma2, every other layer uses sliding window attention. This is the
size of the sliding window.
final_logit_softcapping (`float`, *optional*, defaults to 30.0): scaling factor when applying tanh softcapping on the logits.
attn_logit_softcapping (`float`, *optional*, defaults to 50.0): scaling factor when applying tanh softcapping on the attention scores.
cache_implementation (`str`, *optional*, defaults to `"hybrid"`): the cache type to be used with `generate`.
```python
>>> from transformers import Gemma2Model, Gemma2Config
>>> # Initializing a Gemma2 gemma2-7b style configuration
>>> configuration = Gemma2Config()
>>> # Initializing a model from the gemma2-7b style configuration
>>> model = Gemma2Model(configuration)
>>> # Accessing the model configuration
>>> configuration = model.config
```"""
model_type = "tpu_gemma2"
keys_to_ignore_at_inference = ["past_key_values"]
def __init__(
self,
vocab_size=256000,
hidden_size=3072,
intermediate_size=24576,
num_hidden_layers=28,
num_attention_heads=16,
num_key_value_heads=16,
head_dim=256,
hidden_activation="gelu_pytorch_tanh",
max_position_embeddings=8192,
initializer_range=0.02,
rms_norm_eps=1e-6,
use_cache=True,
pad_token_id=0,
eos_token_id=1,
bos_token_id=2,
tie_word_embeddings=True,
rope_theta=10000.0,
attention_bias=False,
attention_dropout=0.0,
query_pre_attn_scalar=224,
sliding_window=4096,
final_logit_softcapping=30.0,
attn_logit_softcapping=50.0,
cache_implementation="hybrid",
expand_input_ids=False, # Transformers-native PyTorch generation support
expand_input_ids_maxlen=None,
expand_input_ids_vocab_size=None,
expand_input_ids_dict=None,
**kwargs,
):
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,
)
self.vocab_size = vocab_size
self.max_position_embeddings = max_position_embeddings
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.head_dim = head_dim
self.num_key_value_heads = num_key_value_heads
self.initializer_range = initializer_range
self.rms_norm_eps = rms_norm_eps
self.use_cache = use_cache
self.rope_theta = rope_theta
self.attention_bias = attention_bias
self.attention_dropout = attention_dropout
self.hidden_activation = hidden_activation
self.query_pre_attn_scalar = query_pre_attn_scalar
self.sliding_window = sliding_window
self.final_logit_softcapping = final_logit_softcapping
self.attn_logit_softcapping = attn_logit_softcapping
self.cache_implementation = cache_implementation
self.expand_input_ids = expand_input_ids
self.expand_input_ids_maxlen = expand_input_ids_maxlen
self.expand_input_ids_vocab_size = expand_input_ids_vocab_size
self.expand_input_ids_dict = expand_input_ids_dict
|