configuration_nllbllm2vec.py

from typing import Optional, Dict
from transformers import AutoConfig
from transformers.configuration_utils import PretrainedConfig
from transformers.models.llama.configuration_llama import LlamaConfig
from transformers.models.m2m_100.configuration_m2m_100 import M2M100Config

NLLBLLM2VEC_TYPE = "nllb-llm2vec"

DEFAULT_M2M100_CONFIG = {
    "activation_dropout": 0.0,
    "activation_function": "relu",
    "architectures": ["M2M100Encoder"],
    "attention_dropout": 0.1,
    "bos_token_id": 0,
    "d_model": 1024,
    "decoder_attention_heads": 16,
    "decoder_ffn_dim": 4096,
    "decoder_layerdrop": 0,
    "decoder_layers": 12,
    "decoder_start_token_id": 2,
    "dropout": 0.1,
    "encoder_attention_heads": 16,
    "encoder_ffn_dim": 4096,
    "encoder_layerdrop": 0,
    "encoder_layers": 12,
    "eos_token_id": 2,
    "init_std": 0.02,
    "is_encoder_decoder": True,
    "max_position_embeddings": 1024,
    "model_type": "m2m_100",
    "num_hidden_layers": 12,
    "pad_token_id": 1,
    "scale_embedding": True,
    "torch_dtype": "float32",
    "transformers_version": "4.21.0.dev0",
    "use_cache": True,
    "vocab_size": 256206,
    "tokenizer_class": "NllbTokenizer",
    "max_length": 200,
    "_attn_implementation": "flash_attention_2",
}

DEFAULT_LLAMA_CONFIG = {
    "attention_bias": False,
    "attention_dropout": 0,
    "bos_token_id": 128000,
    "eos_token_id": 128001,
    "hidden_act": "silu",
    "hidden_size": 4096,
    "initializer_range": 0.02,
    "intermediate_size": 14336,
    "max_position_embeddings": 8192,
    "model_type": "llama",
    "num_attention_heads": 32,
    "num_hidden_layers": 32,
    "num_key_value_heads": 8,
    "pretraining_tp": 1,
    "rms_norm_eps": 0.00001,
    "rope_scaling": None,
    "rope_theta": 500000,
    "tie_word_embeddings": False,
    "torch_dtype": "bfloat16",
    "transformers_version": "4.40.0.dev0",
    "use_cache": False,
    "vocab_size": 128256,
    "_attn_implementation": "flash_attention_2",
}


class NLLBLLM2VecConfig(PretrainedConfig):
    model_type = "nllb-llm2vec"
    is_composition = False

    def __init__(
        self,
        nllb_config: Dict = DEFAULT_M2M100_CONFIG,
        llm2vec_config: Dict = DEFAULT_LLAMA_CONFIG,
        _attn_implementation="sdpa",
        initializer_range: Optional[float] = None,
        **kwargs,
    ):
        super().__init__(**kwargs)
        self._attn_implementation = _attn_implementation
        self.nllb_config = M2M100Config(**nllb_config)
        self.nllb_config._attn_implementation = _attn_implementation
        self.llm2vec_config = LlamaConfig(**llm2vec_config)
        self.llm2vec_config._attn_implementation = _attn_implementation
        if initializer_range is None:
            self.initializer_range = self.llm2vec_config.initializer_range
        else:
            self.initializer_range = initializer_range
            self.llm2vec_config.initializer_range


AutoConfig.register(NLLBLLM2VEC_TYPE, NLLBLLM2VecConfig)

NLLBLLM2VecConfig.register_for_auto_class()