Merge pull request #1252 from anvie/neftune

add NEFTune optimization
This commit is contained in:
hoshi-hiyouga 2023-10-22 15:59:20 +08:00 committed by GitHub
commit b42a145253
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
5 changed files with 88 additions and 1 deletions

View File

@ -22,6 +22,8 @@ https://github.com/hiyouga/LLaMA-Factory/assets/16256802/6ba60acc-e2e2-4bec-b846
## Changelog
[23/10/21] We supported [NEFTune](https://arxiv.org/abs/2310.05914) optimization . Try `--neftune_noise_alpha` argument to activate NEFTune, e.g., `--neftune_noise_alpha 5`.
[23/09/27] We supported **$S^2$-Attn** proposed by [LongLoRA](https://github.com/dvlab-research/LongLoRA) for the LLaMA models. Try `--shift_attn` argument to enable shift short attention.
[23/09/23] We integrated MMLU, C-Eval and CMMLU benchmarks in this repo. See [this example](#evaluation) to evaluate your models.

View File

@ -22,6 +22,8 @@ https://github.com/hiyouga/LLaMA-Factory/assets/16256802/6ba60acc-e2e2-4bec-b846
## 更新日志
[23/10/21] 我们支持了 [NEFTune](https://arxiv.org/abs/2310.05914) 优化。试试`--neftune_noise_alpha` 参数来激活 NEFTune例如`--neftune_noise_alpha 5`。
[23/09/27] 我们针对 LLaMA 模型支持了 [LongLoRA](https://github.com/dvlab-research/LongLoRA) 提出的 **$S^2$-Attn**。请使用 `--shift_attn` 参数以启用该功能。
[23/09/23] 我们在项目中集成了 MMLU、C-Eval 和 CMMLU 评估集。使用方法请参阅[此示例](#模型评估)。

View File

@ -75,6 +75,10 @@ class FinetuningArguments:
default=0.1,
metadata={"help": "The beta parameter for the DPO loss."}
)
neftune_noise_alpha: Optional[float] = field(
default=None,
metadata={"help": "The alpha parameter for the NEFTune noise. By setting this the NEFTune optimization will be activated."}
)
def __post_init__(self):
if isinstance(self.lora_target, str): # support custom target modules/layers of LoRA

View File

@ -3,8 +3,10 @@ import json
import torch
import numpy as np
import torch.nn as nn
from functools import wraps
from typing import TYPE_CHECKING, Any, Dict, List, Optional, Tuple, Union
from transformers import Seq2SeqTrainer
from transformers import Seq2SeqTrainer, PreTrainedModel, Trainer
from peft import PeftModel
from llmtuner.extras.constants import IGNORE_INDEX
from llmtuner.extras.logging import get_logger
@ -21,6 +23,14 @@ class CustomSeq2SeqTrainer(Seq2SeqTrainer):
Inherits PeftTrainer to compute generative metrics such as BLEU and ROUGE.
"""
def __init__(self, model: Union["PreTrainedModel", nn.Module] = None, neftune_noise_alpha: Optional[float] = 0, **kwargs):
super().__init__(model, **kwargs)
self.neftune_noise_alpha = neftune_noise_alpha
self._neftune_activated = False
if self.neftune_noise_alpha:
self._activate_neftune(model)
def prediction_step(
self,
model: nn.Module,
@ -99,3 +109,71 @@ class CustomSeq2SeqTrainer(Seq2SeqTrainer):
for pred, label in zip(decoded_preds, decoded_labels):
res.append(json.dumps({"label": label, "predict": pred}, ensure_ascii=False))
writer.write("\n".join(res))
@wraps(Trainer.train)
def train(self, *args, **kwargs):
output = super().train(*args, **kwargs)
# After training we make sure to retrieve back the original forward pass method
# for the embedding layer.
if self.neftune_noise_alpha is not None:
self._deactivate_neftune(self.model)
return output
def _toggle_neftune(self, model, activate=True):
"""Toggle NEFTune optimization for a model (i.e. activate or deactivate).
This optimization based on this paper: https://arxiv.org/abs/2310.05914
Parameters:
model : PreTrainedModel or PeftModel
The model to toggle the noise for.
activate : bool, optional (default=True)
Whether to activate the noise or not.
"""
if activate == self._neftune_activated:
return
self._neftune_activated = activate
embeddings = (model.get_input_embeddings() if isinstance(model, PreTrainedModel)
else model.base_model.get_input_embeddings() if isinstance(model, PeftModel)
else None)
if embeddings:
if activate:
embeddings.neftune_noise_alpha = self.neftune_noise_alpha
embeddings._trl_old_forward = embeddings.forward
neftune_method = _neftune_forward_function.__get__(embeddings, embeddings.__class__)
setattr(embeddings, "forward", neftune_method)
logger.info("NEFTune activated with alpha: ", self.neftune_noise_alpha)
elif hasattr(embeddings, "_trl_old_forward"):
embeddings.forward = embeddings._trl_old_forward
del embeddings._trl_old_forward
del embeddings.neftune_noise_alpha
logger.info("NEFTune deactivated")
_activate_neftune = lambda self, model: self._toggle_neftune(model, activate=True)
_deactivate_neftune = lambda self, model: self._toggle_neftune(model, activate=False)
def _neftune_forward_function(self, input: torch.Tensor) -> torch.Tensor:
"""
This code is adapted from the original source code that can be found here: https://github.com/neelsjain/NEFTune
"""
embeddings = torch.nn.functional.embedding(
input,
self.weight,
self.padding_idx,
self.max_norm,
self.norm_type,
self.scale_grad_by_freq,
self.sparse)
if self.training:
dims = torch.tensor(embeddings.size(1) * embeddings.size(2))
mag_norm = self.neftune_noise_alpha / torch.sqrt(dims)
embeddings += torch.zeros_like(embeddings).uniform_(-mag_norm, mag_norm)
return embeddings

View File

@ -53,6 +53,7 @@ def run_sft(
data_collator=data_collator,
callbacks=callbacks,
compute_metrics=ComputeMetrics(tokenizer) if training_args.predict_with_generate else None,
neftune_noise_alpha=finetuning_args.neftune_noise_alpha,
**split_dataset(dataset, data_args, training_args)
)