File size: 4,257 Bytes
9031f04 |
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 |
import json
import os
from typing import TYPE_CHECKING, Any, Dict, List, Optional, Tuple, Union
import numpy as np
import torch
import torch.nn as nn
from transformers import Seq2SeqTrainer
from ...extras.constants import IGNORE_INDEX
from ...extras.logging import get_logger
if TYPE_CHECKING:
from transformers.trainer import PredictionOutput
logger = get_logger(__name__)
class CustomSeq2SeqTrainer(Seq2SeqTrainer):
r"""
Inherits PeftTrainer to compute generative metrics such as BLEU and ROUGE.
"""
def prediction_step(
self,
model: 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 # backup labels
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: # truncate the labels instead of padding the inputs (llama2 fp16 compatibility)
inputs["labels"] = inputs["labels"][:, :prompt_len]
loss, generated_tokens, _ = super().prediction_step( # ignore the returned labels (may be truncated)
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 # adopt left-padding
return padded_tensor.contiguous() # in contiguous memory
def save_predictions(self, 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
) # move pad token to last
decoded_labels = self.tokenizer.batch_decode(
labels, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
decoded_preds = self.tokenizer.batch_decode(preds, skip_special_tokens=True, clean_up_tokenization_spaces=True)
with open(output_prediction_file, "w", encoding="utf-8") as writer:
res: List[str] = []
for label, pred in zip(decoded_labels, decoded_preds):
res.append(json.dumps({"label": label, "predict": pred}, ensure_ascii=False))
writer.write("\n".join(res))
|