| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
|
|
| import json |
| import os |
| from types import MethodType |
| from typing import TYPE_CHECKING, Any, Dict, List, Optional, Tuple, Union |
|
|
| import numpy as np |
| import torch |
| from transformers import Seq2SeqTrainer |
|
|
| from ...extras.constants import IGNORE_INDEX |
| from ...extras.logging import get_logger |
| from ..callbacks import PissaConvertCallback, SaveProcessorCallback |
| from ..trainer_utils import create_custom_optimzer, create_custom_scheduler |
|
|
|
|
| if TYPE_CHECKING: |
| from torch.utils.data import Dataset |
| from transformers import ProcessorMixin |
| from transformers.trainer import PredictionOutput |
|
|
| from ...hparams import FinetuningArguments |
|
|
|
|
| logger = get_logger(__name__) |
|
|
|
|
| class CustomSeq2SeqTrainer(Seq2SeqTrainer): |
| r""" |
| Inherits Seq2SeqTrainer to compute generative metrics such as BLEU and ROUGE. |
| """ |
|
|
| def __init__( |
| self, finetuning_args: "FinetuningArguments", processor: Optional["ProcessorMixin"], **kwargs |
| ) -> None: |
| super().__init__(**kwargs) |
| self.finetuning_args = finetuning_args |
|
|
| if processor is not None: |
| self.add_callback(SaveProcessorCallback(processor)) |
|
|
| if finetuning_args.pissa_convert: |
| self.add_callback(PissaConvertCallback) |
|
|
| if finetuning_args.use_badam: |
| from badam import BAdamCallback, clip_grad_norm_old_version |
|
|
| self.accelerator.clip_grad_norm_ = MethodType(clip_grad_norm_old_version, self.accelerator) |
| self.add_callback(BAdamCallback) |
|
|
| def create_optimizer(self) -> "torch.optim.Optimizer": |
| if self.optimizer is None: |
| self.optimizer = create_custom_optimzer(self.model, self.args, self.finetuning_args) |
| return super().create_optimizer() |
|
|
| def create_scheduler( |
| self, num_training_steps: int, optimizer: Optional["torch.optim.Optimizer"] = None |
| ) -> "torch.optim.lr_scheduler.LRScheduler": |
| create_custom_scheduler(self.args, num_training_steps, optimizer) |
| return super().create_scheduler(num_training_steps, optimizer) |
|
|
| def prediction_step( |
| self, |
| model: "torch.nn.Module", |
| inputs: Dict[str, Union[torch.Tensor, Any]], |
| prediction_loss_only: bool, |
| ignore_keys: Optional[List[str]] = None, |
| ) -> Tuple[Optional[float], Optional[torch.Tensor], Optional[torch.Tensor]]: |
| r""" |
| Removes the prompt part in the generated tokens. |
| |
| Subclass and override to inject custom behavior. |
| """ |
| labels = inputs["labels"].detach().clone() if "labels" in inputs else None |
| if self.args.predict_with_generate: |
| assert self.tokenizer.padding_side == "left", "This method only accepts left-padded tensor." |
| prompt_len, label_len = inputs["input_ids"].size(-1), inputs["labels"].size(-1) |
| if prompt_len > label_len: |
| inputs["labels"] = self._pad_tensors_to_target_len(inputs["labels"], inputs["input_ids"]) |
| if label_len > prompt_len: |
| inputs["labels"] = inputs["labels"][:, :prompt_len] |
|
|
| loss, generated_tokens, _ = super().prediction_step( |
| model, inputs, prediction_loss_only=prediction_loss_only, ignore_keys=ignore_keys |
| ) |
| if generated_tokens is not None and self.args.predict_with_generate: |
| generated_tokens[:, :prompt_len] = self.tokenizer.pad_token_id |
| generated_tokens = generated_tokens.contiguous() |
|
|
| return loss, generated_tokens, labels |
|
|
| def _pad_tensors_to_target_len(self, src_tensor: torch.Tensor, tgt_tensor: torch.Tensor) -> torch.Tensor: |
| r""" |
| Pads the tensor to the same length as the target tensor. |
| """ |
| assert self.tokenizer.pad_token_id is not None, "Pad token is required." |
| padded_tensor = self.tokenizer.pad_token_id * torch.ones_like(tgt_tensor) |
| padded_tensor[:, -src_tensor.shape[-1] :] = src_tensor |
| return padded_tensor.contiguous() |
|
|
| def save_predictions(self, dataset: "Dataset", predict_results: "PredictionOutput") -> None: |
| r""" |
| Saves model predictions to `output_dir`. |
| |
| A custom behavior that not contained in Seq2SeqTrainer. |
| """ |
| if not self.is_world_process_zero(): |
| return |
|
|
| output_prediction_file = os.path.join(self.args.output_dir, "generated_predictions.jsonl") |
| logger.info(f"Saving prediction results to {output_prediction_file}") |
|
|
| labels = np.where( |
| predict_results.label_ids != IGNORE_INDEX, predict_results.label_ids, self.tokenizer.pad_token_id |
| ) |
| preds = np.where( |
| predict_results.predictions != IGNORE_INDEX, predict_results.predictions, self.tokenizer.pad_token_id |
| ) |
|
|
| for i in range(len(preds)): |
| pad_len = np.nonzero(preds[i] != self.tokenizer.pad_token_id)[0] |
| if len(pad_len): |
| preds[i] = np.concatenate((preds[i][pad_len[0] :], preds[i][: pad_len[0]]), axis=-1) |
|
|
| decoded_inputs = self.tokenizer.batch_decode(dataset["input_ids"], skip_special_tokens=True) |
| decoded_labels = self.tokenizer.batch_decode(labels, skip_special_tokens=True) |
| decoded_preds = self.tokenizer.batch_decode(preds, skip_special_tokens=True) |
|
|
| with open(output_prediction_file, "w", encoding="utf-8") as writer: |
| res: List[str] = [] |
| for text, label, pred in zip(decoded_inputs, decoded_labels, decoded_preds): |
| res.append(json.dumps({"prompt": text, "label": label, "predict": pred}, ensure_ascii=False)) |
|
|
| writer.write("\n".join(res)) |
|
|
