wath5/kgl_lmsys_pref_classif

Model Card for Model ID

Given a (Query, ModelAAnswer, ModelBAnswer) This model gives a vector in 3D like lMSYS (ModelAWin Proba), (ModelBWin Proba), (Tie Proba)

Model Details

Model Description

This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.

• Developed by: @sayoulala (Yang Zhou)

• Model type: Gemma for Sentence Classification

• Language(s) (NLP): English Only

Model Sources [optional]

• Repository: [More Information Needed]
• Paper [optional]: [More Information Needed]
• Demo [optional]: [More Information Needed]

Uses

Mimic human preference given a query and 2 different answers.

Direct Use

import torch
from torch import nn
from torch.nn import CrossEntropyLoss, MSELoss, BCEWithLogitsLoss
from transformers import Gemma2PreTrainedModel,Gemma2Model, Cache, AutoTokenizer
from transformers.modeling_outputs import SequenceClassifierOutputWithPast
from typing import Optional, List, Union, Tuple
from dataclasses import dataclass

@dataclass
class Config:
    gemma_dir = 'wath5/kgl_lmsys_pref_classif'
    max_length = 2000
    batch_size = 8
    device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")

cfg = Config()

class Gemma2ForSequenceClassificationV1(Gemma2PreTrainedModel):
    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels
        self.model = Gemma2Model(config)
        self.score = nn.Linear(config.hidden_size, self.num_labels, bias=False)

        # Initialize weights and apply final processing
        self.post_init()

    def get_input_embeddings(self):
        return self.model.embed_tokens

    def set_input_embeddings(self, value):
        self.model.embed_tokens = value

    def forward(
            self,
            input_ids: torch.LongTensor = None,
            attention_mask: Optional[torch.Tensor] = None,
            position_ids: Optional[torch.LongTensor] = None,
            past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None,
            inputs_embeds: Optional[torch.FloatTensor] = None,
            labels: Optional[torch.LongTensor] = None,
            use_cache: Optional[bool] = None,
            output_attentions: Optional[bool] = None,
            output_hidden_states: Optional[bool] = None,
            return_dict: Optional[bool] = None,
    ) -> Union[Tuple, SequenceClassifierOutputWithPast]:
        r"""
        labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
            config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
            `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        transformer_outputs = self.model(
            input_ids,
            attention_mask=attention_mask,
            position_ids=position_ids,
            past_key_values=past_key_values,
            inputs_embeds=inputs_embeds,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
        )
        hidden_states = transformer_outputs[0]
#         logits = self.score(hidden_states)

        if input_ids is not None:
            batch_size = input_ids.shape[0]
        else:
            batch_size = inputs_embeds.shape[0]

        if self.config.pad_token_id is None and batch_size != 1:
            raise ValueError("Cannot handle batch sizes > 1 if no padding token is defined.")
        if self.config.pad_token_id is None:
            sequence_lengths = -1
        else:
            if input_ids is not None:
                # if no pad token found, use modulo instead of reverse indexing for ONNX compatibility
                sequence_lengths = torch.eq(input_ids, self.config.pad_token_id).int().argmax(-1) - 1
                sequence_lengths = sequence_lengths % input_ids.shape[-1]
                sequence_lengths = sequence_lengths.to(hidden_states.device)
            else:
                sequence_lengths = -1
        hidden_states = hidden_states[
            torch.arange(batch_size, device=hidden_states.device), sequence_lengths]  # eos
        pooled_logits = self.score(hidden_states)
        
        return pooled_logits


tokenizer = AutoTokenizer.from_pretrained("/kaggle/input/v7-dpo-16bit-01234-8bit-all/v7_dpo_16bit_01234_8bit_all")

model = Gemma2ForSequenceClassificationV1.from_pretrained(
    cfg.gemma_dir,
    num_labels=3,
    device_map=cfg.device,
    use_cache=False,
)
model.config.pad_token_id = tokenizer.pad_token_id

How to Get Started with the Model

from transformers.data.data_collator import pad_without_fast_tokenizer_warning

@torch.no_grad()
def single_prompt_inference(prompt, model, device, max_length=cfg.max_length):
    """
    Perform inference on a single prompt.

    Args:
        prompt (str): The input prompt for inference.
        model (torch.nn.Module): The model used for inference.
        device (torch.device): The device to run inference on.
        tokenizer (Tokenizer): Tokenizer for preprocessing input text.
        max_length (int): Maximum sequence length for tokenization.

    Returns:
        dict: Probabilities for "a_win", "b_win", and "tie".
    """
    # Tokenize the input prompt
    input_ids = tokenizer(prompt, truncation=True, max_length=max_length)['input_ids']
    input_ids.append(tokenizer.eos_token_id)  # Add EOS token if needed

    # Prepare inputs
    inputs = pad_without_fast_tokenizer_warning(
        tokenizer,
        {"input_ids": [input_ids]},  # Wrap in a list for compatibility
        padding="max_length",
        pad_to_multiple_of=None,
        max_length=max_length,
        return_tensors="pt",
    )

    # Move inputs to the appropriate device
    inputs = inputs.to(device)

    # Run the model
    outputs = model(**inputs)

    # Get probabilities using softmax
    proba = outputs.softmax(-1).cpu().squeeze()

    return {
        "winner_model_a": proba[0].item(),
        "winner_model_b": proba[1].item(),
        "tie": proba[2].item(),
    }


def create_rounds(query: str,
                  answer_a: str,
                  answer_b: str) -> str:
    prompt =f"""User question:
    \"""{query}\"""
    Answer A:
    \"""{answer_a}\"""
    Answer B:
    \"""{answer_b}\"""
    """
    return prompt

query = "Hello, what is the height of the reassembled blind product?"
answer_a = "Vous pouvez trouver toutes les informations techniques, y compris la hauteur du produit store remonté, directement sur la fiche produit de notre site. Cliquez sur l'onglet 'Produits' dnas la barre de navigation ou utilisez le moteur de recherche pour accéder au produit recherché. Avez vous une autre question ?"
answer_b = "The height of the aluminum Venetian blind is 130 cm."
prompt_direct = create_rounds(query, answer_a, answer_b)

single_prompt_inference(prompt_direct, model=model, device=cfg.device)

Training Details

https://github.com/shyoulala/LMSYS_BlackPearl