训练概述
为何微调？
Cross Encoder 模型通常在 检索和重排序 搜索堆栈中用作第二阶段重排序器。在这种情况下，Cross Encoder 会对检索器（可以是 Sentence Transformer 模型）返回的前 X 个候选项进行重排序。为了避免重排序器模型降低您用例的性能，对其进行微调至关重要。重排序器始终只有一个输出标签。
除此之外，Cross Encoder 模型还可以用作对分类器。例如，在自然语言推理数据上训练的模型可以用于将文本对分类为“矛盾”、“蕴含”和“中性”。对分类器通常具有多个输出标签。
请参阅 训练示例，其中包含许多常见实际应用的训练脚本，您可以采纳这些脚本。
训练组件
训练 Cross Encoder 模型涉及 3 到 5 个组件，就像 训练 Sentence Transformer 模型 一样
数据集
CrossEncoderTrainer 使用 datasets.Dataset（一个数据集）或 datasets.DatasetDict 实例（多个数据集，另请参阅 多数据集训练）进行训练和评估。
如果您想从 Hugging Face Datasets 加载数据，那么您应该使用 datasets.load_dataset()
from datasets import load_dataset
train_dataset = load_dataset("sentence-transformers/all-nli", "pair-class", split="train")
eval_dataset = load_dataset("sentence-transformers/all-nli", "pair-class", split="dev")
print(train_dataset)
"""
Dataset({
features: ['premise', 'hypothesis', 'label'],
num_rows: 942069
})
"""
某些数据集（包括 sentence-transformers/all-nli）要求您在数据集名称旁边提供“subset”。sentence-transformers/all-nli 有 4 个子集，每个子集具有不同的数据格式：pair、pair-class、pair-score、triplet。
注意
许多可以直接与 Sentence Transformers 一起使用的 Hugging Face 数据集都已标记为 sentence-transformers，使您可以通过浏览 https://huggingface.co/datasets?other=sentence-transformers 轻松找到它们。我们强烈建议您浏览这些数据集，以找到可能对您的任务有用的训练数据集。
如果您有常见文件格式的本地数据，那么您可以使用 datasets.load_dataset() 轻松加载这些数据
from datasets import load_dataset
dataset = load_dataset("csv", data_files="my_file.csv")
或
from datasets import load_dataset
dataset = load_dataset("json", data_files="my_file.json")
如果您有需要一些额外预处理的本地数据，我建议您使用 datasets.Dataset.from_dict() 和列表字典来初始化您的数据集，如下所示
from datasets import Dataset
anchors = []
positives = []
# Open a file, do preprocessing, filtering, cleaning, etc.
# and append to the lists
dataset = Dataset.from_dict({
"anchor": anchors,
"positive": positives,
})
字典中的每个键将成为结果数据集中的一列。
数据集格式
重要的是，您的数据集格式与您的损失函数相匹配（或者您选择与您的数据集格式和模型相匹配的损失函数）。验证数据集格式和模型是否与损失函数一起工作涉及三个步骤
根据 损失概述 表，所有未命名为“label”、“labels”、“score”或“scores”的列都被视为输入。剩余列的数量必须与您选择的损失的有效输入数量相匹配。这些列的名称是不相关的，只有顺序很重要。
如果您的损失函数根据 损失概述 表需要标签，那么您的数据集必须具有名为“label”、“labels”、“score”或“scores”的列。此列将自动作为标签。
模型输出标签的数量与根据 损失概述 表的损失所需数量相匹配。
例如，给定一个数据集，其列为 ["text1", "text2", "label"]，其中“label”列具有从 0 到 1 的浮点相似度分数，并且模型输出 1 个标签，我们可以将其与 BinaryCrossEntropyLoss 一起使用，因为
数据集具有此损失函数所需的“label”列。
数据集具有 2 个非标签列，正好是此损失函数所需的数量。
模型具有 1 个输出标签，正好是此损失函数所需的数量。
如果您的列排序不正确，请务必使用 Dataset.select_columns 重新排序您的数据集列。例如，如果您的数据集具有 ["good_answer", "bad_answer", "question"] 列，那么此数据集在技术上可以与需要 (anchor, positive, negative) 三元组的损失一起使用，但 good_answer 列将被视为 anchor，bad_answer 将被视为 positive，而 question 将被视为 negative。
此外，如果您的数据集具有无关的列（例如 sample_id、metadata、source、type），您应该使用 Dataset.remove_columns 删除这些列，否则它们将被用作输入。您也可以使用 Dataset.select_columns 仅保留所需的列。
Hard Negatives Mining
训练 CrossEncoder 模型的成功通常取决于 negatives 的质量，即查询-negative 分数应较低的段落。Negatives 可以分为两种类型
Soft negatives：完全不相关的段落。
Hard negatives：看起来可能与查询相关，但实际上并非相关的段落。
一个简洁的例子是
查询：苹果公司在哪里成立？
Soft Negative：卡奇河大桥是一座帕克小马桁架桥，横跨阿肯色州核桃岭和帕拉古尔德之间的卡奇河。
Hard Negative：富士苹果是一种苹果品种，于 1930 年代后期开发，并于 1962 年投放市场。
最强大的 CrossEncoder 模型通常经过训练以识别 hard negatives，因此能够“挖掘”hard negatives 非常有价值。Sentence Transformers 支持强大的 mine_hard_negatives() 函数，可以在给定查询-答案对数据集的情况下提供帮助
from datasets import load_dataset
from sentence_transformers import SentenceTransformer
from sentence_transformers.util import mine_hard_negatives
# Load the GooAQ dataset: https://huggingface.co/datasets/sentence-transformers/gooaq
train_dataset = load_dataset("sentence-transformers/gooaq", split=f"train").select(range(100_000))
print(train_dataset)
# Mine hard negatives using a very efficient embedding model
embedding_model = SentenceTransformer("sentence-transformers/static-retrieval-mrl-en-v1", device="cpu")
hard_train_dataset = mine_hard_negatives(
train_dataset,
embedding_model,
num_negatives=5, # How many negatives per question-answer pair
range_min=10, # Skip the x most similar samples
range_max=100, # Consider only the x most similar samples
max_score=0.8, # Only consider samples with a similarity score of at most x
absolute_margin=0.1, # Anchor-negative similarity is at least x lower than anchor-positive similarity
relative_margin=0.1, # Anchor-negative similarity is at most 1-x times the anchor-positive similarity, e.g. 90%
sampling_strategy="top", # Sample the top negatives from the range
batch_size=4096, # Use a batch size of 4096 for the embedding model
output_format="labeled-pair", # The output format is (query, passage, label), as required by BinaryCrossEntropyLoss
use_faiss=True, # Using FAISS is recommended to keep memory usage low (pip install faiss-gpu or pip install faiss-cpu)
)
print(hard_train_dataset)
print(hard_train_dataset[1])
点击查看此脚本的输出。
Dataset({
features: ['question', 'answer'],
num_rows: 100000
})
Batches: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 22/22 [00:01<00:00, 12.74it/s]
Batches: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████| 25/25 [00:00<00:00, 37.50it/s]
Querying FAISS index: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████| 7/7 [00:18<00:00, 2.66s/it]
Metric Positive Negative Difference
Count 100,000 436,925
Mean 0.5882 0.4040 0.2157
Median 0.5989 0.4024 0.1836
Std 0.1425 0.0905 0.1013
Min -0.0514 0.1405 0.1014
25% 0.4993 0.3377 0.1352
50% 0.5989 0.4024 0.1836
75% 0.6888 0.4681 0.2699
Max 0.9748 0.7486 0.7545
Skipped 2,420,871 potential negatives (23.97%) due to the absolute_margin of 0.1.
Skipped 43 potential negatives (0.00%) due to the max_score of 0.8.
Could not find enough negatives for 63075 samples (12.62%). Consider adjusting the range_max, range_min, absolute_margin, relative_margin and max_score parameters if you'd like to find more valid negatives.
Dataset({
features: ['question', 'answer', 'label'],
num_rows: 536925
})
{
'question': 'how to transfer bookmarks from one laptop to another?',
'answer': 'Using an External Drive Just about any external drive, including a USB thumb drive, or an SD card can be used to transfer your files from one laptop to another. Connect the drive to your old laptop; drag your files to the drive, then disconnect it and transfer the drive contents onto your new laptop.',
'label': 0
}
损失函数
损失函数量化模型对于给定数据批次的表现，允许优化器更新模型权重以产生更有利（即更低）的损失值。这是训练过程的核心。
遗憾的是，没有一个损失函数适用于所有用例。相反，使用哪个损失函数在很大程度上取决于您的可用数据和您的目标任务。请参阅 数据集格式 以了解哪些数据集对哪些损失函数有效。此外，损失概述 将是您了解选项的最佳朋友。
大多数损失函数都可以仅使用您正在训练的 CrossEncoder 以及一些可选参数进行初始化，例如
from datasets import load_dataset
from sentence_transformers import CrossEncoder
from sentence_transformers.cross_encoder.losses import MultipleNegativesRankingLoss
# Load a model to train/finetune
model = CrossEncoder("xlm-roberta-base", num_labels=1) # num_labels=1 is for rerankers
# Initialize the MultipleNegativesRankingLoss
# This loss requires pairs of related texts or triplets
loss = MultipleNegativesRankingLoss(model)
# Load an example training dataset that works with our loss function:
train_dataset = load_dataset("sentence-transformers/gooaq", split="train")
训练参数
CrossEncoderTrainingArguments 类可用于指定影响训练性能的参数以及定义跟踪/调试参数。虽然它是可选的，但强烈建议尝试各种有用的参数。
learning_rate lr_scheduler_type warmup_ratio num_train_epochs max_steps per_device_train_batch_size per_device_eval_batch_size auto_find_batch_size fp16 bf16 load_best_model_at_end metric_for_best_model gradient_accumulation_steps gradient_checkpointing eval_accumulation_steps optim dataloader_num_workers dataloader_prefetch_factor batch_sampler multi_dataset_batch_sampler以下是如何初始化 CrossEncoderTrainingArguments 的示例
from sentence_transformers.cross_encoder import CrossEncoderTrainingArguments
args = CrossEncoderTrainingArguments(
# Required parameter:
output_dir="models/reranker-MiniLM-msmarco-v1",
# Optional training parameters:
num_train_epochs=1,
per_device_train_batch_size=16,
per_device_eval_batch_size=16,
learning_rate=2e-5,
warmup_ratio=0.1,
fp16=True, # Set to False if you get an error that your GPU can't run on FP16
bf16=False, # Set to True if you have a GPU that supports BF16
batch_sampler=BatchSamplers.NO_DUPLICATES, # losses that use "in-batch negatives" benefit from no duplicates
# Optional tracking/debugging parameters:
eval_strategy="steps",
eval_steps=100,
save_strategy="steps",
save_steps=100,
save_total_limit=2,
logging_steps=100,
run_name="reranker-MiniLM-msmarco-v1", # Will be used in W&B if `wandb` is installed
)
评估器
您可以为 CrossEncoderTrainer 提供 eval_dataset 以在训练期间获得评估损失，但获取训练期间更具体的指标也可能很有用。为此，您可以使用评估器在训练之前、期间或之后使用有用的指标评估模型的性能。您可以同时使用 eval_dataset 和评估器，或者只使用其中一个，或者都不使用。它们基于 eval_strategy 和 eval_steps 训练参数 进行评估。
以下是 Sentence Transformers 附带的已实现的评估器
评估器 |
所需数据 |
|---|---|
带有类标签（二元或多类）的对。 |
|
带有相似度分数的对。 |
|
无需数据。 |
|
|
此外，应使用 SequentialEvaluator 将多个评估器组合到一个评估器中，该评估器可以传递给 CrossEncoderTrainer。
有时您没有所需的评估数据来自己准备这些评估器之一，但您仍然想跟踪模型在一些常见基准上的表现。在这种情况下，您可以将这些评估器与来自 Hugging Face 的数据一起使用。
from sentence_transformers import CrossEncoder
from sentence_transformers.cross_encoder.evaluation import CrossEncoderNanoBEIREvaluator
# Load a model
model = CrossEncoder("cross-encoder/ms-marco-MiniLM-L6-v2")
# Initialize the evaluator. Unlike most other evaluators, this one loads the relevant datasets
# directly from Hugging Face, so there's no mandatory arguments
dev_evaluator = CrossEncoderNanoBEIREvaluator()
# You can run evaluation like so:
# results = dev_evaluator(model)
准备 CrossEncoderRerankingEvaluator 的数据可能很困难，因为除了查询-positive 数据外，您还需要 negatives。
mine_hard_negatives() 函数有一个方便的 include_positives 参数，可以将其设置为 True 以同时挖掘 positive 文本。当作为 documents（必须是 1. 排序的，并且 2. 包含 positives）提供给 CrossEncoderRerankingEvaluator 时，评估器不仅会评估 CrossEncoder 的重排序性能，还会评估用于挖掘的嵌入模型的原始排序。
例如
CrossEncoderRerankingEvaluator: Evaluating the model on the gooaq-dev dataset:
Queries: 1000 Positives: Min 1.0, Mean 1.0, Max 1.0 Negatives: Min 49.0, Mean 49.1, Max 50.0
Base -> Reranked
MAP: 53.28 -> 67.28
MRR@10: 52.40 -> 66.65
NDCG@10: 59.12 -> 71.35
请注意，默认情况下，如果您将 CrossEncoderRerankingEvaluator 与 documents 一起使用，评估器将使用所有 positives 进行重排序，即使它们不在文档中。这对于从您的评估器中获得更强的信号很有用，但确实给出了略微不切实际的性能。毕竟，最大性能现在是 100，而通常情况下，它受第一阶段检索器是否实际检索到 positives 的限制。
您可以通过在初始化 CrossEncoderRerankingEvaluator 时设置 always_rerank_positives=False 来启用真实的行为。使用此真实的二阶段性能重复相同的脚本会导致
CrossEncoderRerankingEvaluator: Evaluating the model on the gooaq-dev dataset:
Queries: 1000 Positives: Min 1.0, Mean 1.0, Max 1.0 Negatives: Min 49.0, Mean 49.1, Max 50.0
Base -> Reranked
MAP: 53.28 -> 66.12
MRR@10: 52.40 -> 65.61
NDCG@10: 59.12 -> 70.10
from datasets import load_dataset
from sentence_transformers import SentenceTransformer
from sentence_transformers.cross_encoder import CrossEncoder
from sentence_transformers.cross_encoder.evaluation import CrossEncoderRerankingEvaluator
from sentence_transformers.util import mine_hard_negatives
# Load a model
model = CrossEncoder("cross-encoder/ms-marco-MiniLM-L6-v2")
# Load the GooAQ dataset: https://huggingface.co/datasets/sentence-transformers/gooaq
full_dataset = load_dataset("sentence-transformers/gooaq", split=f"train").select(range(100_000))
dataset_dict = full_dataset.train_test_split(test_size=1_000, seed=12)
train_dataset = dataset_dict["train"]
eval_dataset = dataset_dict["test"]
print(eval_dataset)
"""
Dataset({
features: ['question', 'answer'],
num_rows: 1000
})
"""
# Mine hard negatives using a very efficient embedding model
embedding_model = SentenceTransformer("sentence-transformers/static-retrieval-mrl-en-v1", device="cpu")
hard_eval_dataset = mine_hard_negatives(
eval_dataset,
embedding_model,
corpus=full_dataset["answer"], # Use the full dataset as the corpus
num_negatives=50, # How many negatives per question-answer pair
batch_size=4096, # Use a batch size of 4096 for the embedding model
output_format="n-tuple", # The output format is (query, positive, negative1, negative2, ...) for the evaluator
include_positives=True, # Key: Include the positive answer in the list of negatives
use_faiss=True, # Using FAISS is recommended to keep memory usage low (pip install faiss-gpu or pip install faiss-cpu)
)
print(hard_eval_dataset)
"""
Dataset({
features: ['question', 'answer', 'negative_1', 'negative_2', 'negative_3', 'negative_4', 'negative_5', 'negative_6', 'negative_7', 'negative_8', 'negative_9', 'negative_10', 'negative_11', 'negative_12', 'negative_13', 'negative_14', 'negative_15', 'negative_16', 'negative_17', 'negative_18', 'negative_19', 'negative_20', 'negative_21', 'negative_22', 'negative_23', 'negative_24', 'negative_25', 'negative_26', 'negative_27', 'negative_28', 'negative_29', 'negative_30', 'negative_31', 'negative_32', 'negative_33', 'negative_34', 'negative_35', 'negative_36', 'negative_37', 'negative_38', 'negative_39', 'negative_40', 'negative_41', 'negative_42', 'negative_43', 'negative_44', 'negative_45', 'negative_46', 'negative_47', 'negative_48', 'negative_49', 'negative_50'],
num_rows: 1000
})
"""
reranking_evaluator = CrossEncoderRerankingEvaluator(
samples=[
{
"query": sample["question"],
"positive": [sample["answer"]],
"documents": [sample[column_name] for column_name in hard_eval_dataset.column_names[2:]],
}
for sample in hard_eval_dataset
],
batch_size=32,
name="gooaq-dev",
)
# You can run evaluation like so
results = reranking_evaluator(model)
"""
CrossEncoderRerankingEvaluator: Evaluating the model on the gooaq-dev dataset:
Queries: 1000 Positives: Min 1.0, Mean 1.0, Max 1.0 Negatives: Min 49.0, Mean 49.1, Max 50.0
Base -> Reranked
MAP: 53.28 -> 67.28
MRR@10: 52.40 -> 66.65
NDCG@10: 59.12 -> 71.35
"""
# {'gooaq-dev_map': 0.6728370126462222, 'gooaq-dev_mrr@10': 0.6665190476190477, 'gooaq-dev_ndcg@10': 0.7135068904582963, 'gooaq-dev_base_map': 0.5327714512001362, 'gooaq-dev_base_mrr@10': 0.5239674603174603, 'gooaq-dev_base_ndcg@10': 0.5912299141913905}
from datasets import load_dataset
from sentence_transformers import CrossEncoder
from sentence_transformers.cross_encoder.evaluation import CrossEncoderCorrelationEvaluator
# Load a model
model = CrossEncoder("cross-encoder/stsb-TinyBERT-L4")
# Load the STSB dataset (https://huggingface.co/datasets/sentence-transformers/stsb)
eval_dataset = load_dataset("sentence-transformers/stsb", split="validation")
pairs = list(zip(eval_dataset["sentence1"], eval_dataset["sentence2"]))
# Initialize the evaluator
dev_evaluator = CrossEncoderCorrelationEvaluator(
sentence_pairs=pairs,
scores=eval_dataset["score"],
name="sts_dev",
)
# You can run evaluation like so:
# results = dev_evaluator(model)
from datasets import load_dataset
from sentence_transformers import CrossEncoder
from sentence_transformers.evaluation import TripletEvaluator, SimilarityFunction
# Load a model
model = CrossEncoder("cross-encoder/nli-deberta-v3-base")
# Load triplets from the AllNLI dataset (https://huggingface.co/datasets/sentence-transformers/all-nli)
max_samples = 1000
eval_dataset = load_dataset("sentence-transformers/all-nli", "pair-class", split=f"dev[:{max_samples}]")
# Create a list of pairs, and map the labels to the labels that the model knows
pairs = list(zip(eval_dataset["premise"], eval_dataset["hypothesis"]))
label_mapping = {0: 1, 1: 2, 2: 0}
labels = [label_mapping[label] for label in eval_dataset["label"]]
# Initialize the evaluator
cls_evaluator = CrossEncoderClassificationEvaluator(
sentence_pairs=pairs,
labels=labels,
name="all-nli-dev",
)
# You can run evaluation like so:
# results = cls_evaluator(model)
警告
当使用 分布式训练 时，评估器仅在第一个设备上运行，这与在所有设备之间共享的训练和评估数据集不同。
训练器
CrossEncoderTrainer 是所有先前组件汇集在一起的地方。我们只需要使用模型、训练参数（可选）、训练数据集、评估数据集（可选）、损失函数、评估器（可选）指定训练器，我们就可以开始训练。让我们看一下将所有这些组件组合在一起的脚本
import logging
import traceback
from datasets import load_dataset
from sentence_transformers.cross_encoder import (
CrossEncoder,
CrossEncoderModelCardData,
CrossEncoderTrainer,
CrossEncoderTrainingArguments,
)
from sentence_transformers.cross_encoder.evaluation import CrossEncoderNanoBEIREvaluator
from sentence_transformers.cross_encoder.losses import CachedMultipleNegativesRankingLoss
# Set the log level to INFO to get more information
logging.basicConfig(format="%(asctime)s - %(message)s", datefmt="%Y-%m-%d %H:%M:%S", level=logging.INFO)
model_name = "microsoft/MiniLM-L12-H384-uncased"
train_batch_size = 64
num_epochs = 1
num_rand_negatives = 5 # How many random negatives should be used for each question-answer pair
# 1a. Load a model to finetune with 1b. (Optional) model card data
model = CrossEncoder(
model_name,
model_card_data=CrossEncoderModelCardData(
language="en",
license="apache-2.0",
model_name="MiniLM-L12-H384 trained on GooAQ",
),
)
print("Model max length:", model.max_length)
print("Model num labels:", model.num_labels)
# 2. Load the GooAQ dataset: https://huggingface.co/datasets/sentence-transformers/gooaq
logging.info("Read the gooaq training dataset")
full_dataset = load_dataset("sentence-transformers/gooaq", split="train").select(range(100_000))
dataset_dict = full_dataset.train_test_split(test_size=1_000, seed=12)
train_dataset = dataset_dict["train"]
eval_dataset = dataset_dict["test"]
logging.info(train_dataset)
logging.info(eval_dataset)
# 3. Define our training loss.
loss = CachedMultipleNegativesRankingLoss(
model=model,
num_negatives=num_rand_negatives,
mini_batch_size=32, # Informs the memory usage
)
# 4. Use CrossEncoderNanoBEIREvaluator, a light-weight evaluator for English reranking
evaluator = CrossEncoderNanoBEIREvaluator(
dataset_names=["msmarco", "nfcorpus", "nq"],
batch_size=train_batch_size,
)
evaluator(model)
# 5. Define the training arguments
short_model_name = model_name if "/" not in model_name else model_name.split("/")[-1]
run_name = f"reranker-{short_model_name}-gooaq-cmnrl"
args = CrossEncoderTrainingArguments(
# Required parameter:
output_dir=f"models/{run_name}",
# Optional training parameters:
num_train_epochs=num_epochs,
per_device_train_batch_size=train_batch_size,
per_device_eval_batch_size=train_batch_size,
learning_rate=2e-5,
warmup_ratio=0.1,
fp16=False, # Set to False if you get an error that your GPU can't run on FP16
bf16=True, # Set to True if you have a GPU that supports BF16
# Optional tracking/debugging parameters:
eval_strategy="steps",
eval_steps=100,
save_strategy="steps",
save_steps=100,
save_total_limit=2,
logging_steps=50,
logging_first_step=True,
run_name=run_name, # Will be used in W&B if `wandb` is installed
seed=12,
)
# 6. Create the trainer & start training
trainer = CrossEncoderTrainer(
model=model,
args=args,
train_dataset=train_dataset,
eval_dataset=eval_dataset,
loss=loss,
evaluator=evaluator,
)
trainer.train()
# 7. Evaluate the final model, useful to include these in the model card
evaluator(model)
# 8. Save the final model
final_output_dir = f"models/{run_name}/final"
model.save_pretrained(final_output_dir)
# 9. (Optional) save the model to the Hugging Face Hub!
# It is recommended to run `huggingface-cli login` to log into your Hugging Face account first
try:
model.push_to_hub(run_name)
except Exception:
logging.error(
f"Error uploading model to the Hugging Face Hub:\n{traceback.format_exc()}To upload it manually, you can run "
f"`huggingface-cli login`, followed by loading the model using `model = CrossEncoder({final_output_dir!r})` "
f"and saving it using `model.push_to_hub('{run_name}')`."
)
import logging
import traceback
import torch
from datasets import load_dataset
from sentence_transformers import SentenceTransformer
from sentence_transformers.cross_encoder import (
CrossEncoder,
CrossEncoderModelCardData,
CrossEncoderTrainer,
CrossEncoderTrainingArguments,
)
from sentence_transformers.cross_encoder.evaluation import (
CrossEncoderNanoBEIREvaluator,
CrossEncoderRerankingEvaluator,
)
from sentence_transformers.cross_encoder.losses import BinaryCrossEntropyLoss
from sentence_transformers.evaluation import SequentialEvaluator
from sentence_transformers.util import mine_hard_negatives
# Set the log level to INFO to get more information
logging.basicConfig(format="%(asctime)s - %(message)s", datefmt="%Y-%m-%d %H:%M:%S", level=logging.INFO)
def main():
model_name = "answerdotai/ModernBERT-base"
train_batch_size = 64
num_epochs = 1
num_hard_negatives = 5 # How many hard negatives should be mined for each question-answer pair
# 1a. Load a model to finetune with 1b. (Optional) model card data
model = CrossEncoder(
model_name,
model_card_data=CrossEncoderModelCardData(
language="en",
license="apache-2.0",
model_name="ModernBERT-base trained on GooAQ",
),
)
print("Model max length:", model.max_length)
print("Model num labels:", model.num_labels)
# 2a. Load the GooAQ dataset: https://huggingface.co/datasets/sentence-transformers/gooaq
logging.info("Read the gooaq training dataset")
full_dataset = load_dataset("sentence-transformers/gooaq", split="train").select(range(100_000))
dataset_dict = full_dataset.train_test_split(test_size=1_000, seed=12)
train_dataset = dataset_dict["train"]
eval_dataset = dataset_dict["test"]
logging.info(train_dataset)
logging.info(eval_dataset)
# 2b. Modify our training dataset to include hard negatives using a very efficient embedding model
embedding_model = SentenceTransformer("sentence-transformers/static-retrieval-mrl-en-v1", device="cpu")
hard_train_dataset = mine_hard_negatives(
train_dataset,
embedding_model,
num_negatives=num_hard_negatives, # How many negatives per question-answer pair
margin=0, # Similarity between query and negative samples should be x lower than query-positive similarity
range_min=0, # Skip the x most similar samples
range_max=100, # Consider only the x most similar samples
sampling_strategy="top", # Sample the top negatives from the range
batch_size=4096, # Use a batch size of 4096 for the embedding model
output_format="labeled-pair", # The output format is (query, passage, label), as required by BinaryCrossEntropyLoss
use_faiss=True,
)
logging.info(hard_train_dataset)
# 2c. (Optionally) Save the hard training dataset to disk
# hard_train_dataset.save_to_disk("gooaq-hard-train")
# Load again with:
# hard_train_dataset = load_from_disk("gooaq-hard-train")
# 3. Define our training loss.
# pos_weight is recommended to be set as the ratio between positives to negatives, a.k.a. `num_hard_negatives`
loss = BinaryCrossEntropyLoss(model=model, pos_weight=torch.tensor(num_hard_negatives))
# 4a. Define evaluators. We use the CrossEncoderNanoBEIREvaluator, which is a light-weight evaluator for English reranking
nano_beir_evaluator = CrossEncoderNanoBEIREvaluator(
dataset_names=["msmarco", "nfcorpus", "nq"],
batch_size=train_batch_size,
)
# 4b. Define a reranking evaluator by mining hard negatives given query-answer pairs
# We include the positive answer in the list of negatives, so the evaluator can use the performance of the
# embedding model as a baseline.
hard_eval_dataset = mine_hard_negatives(
eval_dataset,
embedding_model,
corpus=full_dataset["answer"], # Use the full dataset as the corpus
num_negatives=30, # How many documents to rerank
batch_size=4096,
include_positives=True,
output_format="n-tuple",
use_faiss=True,
)
logging.info(hard_eval_dataset)
reranking_evaluator = CrossEncoderRerankingEvaluator(
samples=[
{
"query": sample["question"],
"positive": [sample["answer"]],
"documents": [sample[column_name] for column_name in hard_eval_dataset.column_names[2:]],
}
for sample in hard_eval_dataset
],
batch_size=train_batch_size,
name="gooaq-dev",
# Realistic setting: only rerank the positives that the retriever found
# Set to True to rerank *all* positives
always_rerank_positives=False,
)
# 4c. Combine the evaluators & run the base model on them
evaluator = SequentialEvaluator([reranking_evaluator, nano_beir_evaluator])
evaluator(model)
# 5. Define the training arguments
short_model_name = model_name if "/" not in model_name else model_name.split("/")[-1]
run_name = f"reranker-{short_model_name}-gooaq-bce"
args = CrossEncoderTrainingArguments(
# Required parameter:
output_dir=f"models/{run_name}",
# Optional training parameters:
num_train_epochs=num_epochs,
per_device_train_batch_size=train_batch_size,
per_device_eval_batch_size=train_batch_size,
learning_rate=2e-5,
warmup_ratio=0.1,
fp16=False, # Set to False if you get an error that your GPU can't run on FP16
bf16=True, # Set to True if you have a GPU that supports BF16
dataloader_num_workers=4,
load_best_model_at_end=True,
metric_for_best_model="eval_gooaq-dev_ndcg@10",
# Optional tracking/debugging parameters:
eval_strategy="steps",
eval_steps=1000,
save_strategy="steps",
save_steps=1000,
save_total_limit=2,
logging_steps=200,
logging_first_step=True,
run_name=run_name, # Will be used in W&B if `wandb` is installed
seed=12,
)
# 6. Create the trainer & start training
trainer = CrossEncoderTrainer(
model=model,
args=args,
train_dataset=hard_train_dataset,
loss=loss,
evaluator=evaluator,
)
trainer.train()
# 7. Evaluate the final model, useful to include these in the model card
evaluator(model)
# 8. Save the final model
final_output_dir = f"models/{run_name}/final"
model.save_pretrained(final_output_dir)
# 9. (Optional) save the model to the Hugging Face Hub!
# It is recommended to run `huggingface-cli login` to log into your Hugging Face account first
try:
model.push_to_hub(run_name)
except Exception:
logging.error(
f"Error uploading model to the Hugging Face Hub:\n{traceback.format_exc()}To upload it manually, you can run "
f"`huggingface-cli login`, followed by loading the model using `model = CrossEncoder({final_output_dir!r})` "
f"and saving it using `model.push_to_hub('{run_name}')`."
)
if __name__ == "__main__":
main()
回调
此 CrossEncoder 训练器集成了对各种 transformers.TrainerCallback 子类的支持，例如
WandbCallback，以便在安装wandb的情况下自动将训练指标记录到 W&BTensorBoardCallback，以便在可以访问tensorboard的情况下将训练指标记录到 TensorBoard。CodeCarbonCallback，以便在安装codecarbon的情况下跟踪模型在训练期间的碳排放量。注意：这些碳排放量将包含在您自动生成的模型卡中。
请参阅 Transformers Callbacks 文档，以获取有关集成回调以及如何编写自己的回调的更多信息。
多数据集训练
性能最佳的模型通常使用多个数据集同时进行训练。通常，这相当棘手，因为每个数据集都具有不同的格式。但是，CrossEncoderTrainer 可以使用多个数据集进行训练，而无需将每个数据集转换为相同的格式。它甚至可以为每个数据集应用不同的损失函数。使用多个数据集进行训练的步骤是
使用
Dataset实例（或DatasetDict）的字典作为train_dataset（以及可选的eval_dataset）。（可选）使用损失函数字典，将数据集名称映射到损失。仅当您希望为不同的数据集使用不同的损失函数时才需要。
每个训练/评估批次将仅包含来自其中一个数据集的样本。从多个数据集中采样批次的顺序由 MultiDatasetBatchSamplers 枚举定义，该枚举可以通过 multi_dataset_batch_sampler 传递给 CrossEncoderTrainingArguments。有效选项包括
MultiDatasetBatchSamplers.ROUND_ROBIN：从每个数据集进行轮询采样，直到其中一个数据集耗尽。使用此策略，可能并非使用每个数据集中的所有样本，但会公平地从每个数据集中采样。MultiDatasetBatchSamplers.PROPORTIONAL（默认）：根据每个数据集的大小比例从每个数据集中采样。使用此策略，将使用每个数据集中的所有样本，并且从较大的数据集中采样的频率更高。
训练技巧
Cross Encoder 模型有其独特的怪癖，因此这里有一些技巧可以帮助您
CrossEncoder模型过度拟合速度相当快，因此建议使用评估器，例如CrossEncoderNanoBEIREvaluator或CrossEncoderRerankingEvaluator，并结合load_best_model_at_end和metric_for_best_model训练参数，以加载训练后评估性能最佳的模型。CrossEncoder特别容易接受强硬负样本 (mine_hard_negatives())。它们教会模型变得非常严格，例如，在区分回答问题的段落或与问题相关的段落时非常有用。请注意，如果您仅使用硬负样本，您的模型在较简单的任务上可能会意外地表现更差。 这可能意味着，对来自第一阶段检索系统（例如，使用
SentenceTransformer模型）的前 200 个结果进行重排序，实际上可能会比对前 100 个结果进行重排序得到更差的前 10 个结果。与硬负样本一起使用随机负样本进行训练可以缓解这种情况。
不要低估
BinaryCrossEntropyLoss，尽管它比 learning-to-rank (LambdaLoss,ListNetLoss) 或 in-batch negatives (CachedMultipleNegativesRankingLoss,MultipleNegativesRankingLoss) 损失更简单，但它仍然是一个非常强大的选择，并且其数据易于准备，尤其是在使用mine_hard_negatives()的情况下。
已弃用的训练
在 Sentence Transformers v4.0 版本之前，模型将使用 CrossEncoder.fit() 方法和 DataLoader 的 InputExample 进行训练，如下所示
from sentence_transformers import CrossEncoder, InputExample
from torch.utils.data import DataLoader
# Define the model. Either from scratch of by loading a pre-trained model
model = CrossEncoder("distilbert/distilbert-base-uncased")
# Define your train examples. You need more than just two examples...
train_examples = [
InputExample(texts=["What are pandas?", "The giant panda ..."], label=1),
InputExample(texts=["What's a panda?", "Mount Vesuvius is a ..."], label=0),
]
# Define your train dataset, the dataloader and the train loss
train_dataloader = DataLoader(train_examples, shuffle=True, batch_size=16)
# Tune the model
model.fit(train_dataloader=train_dataloader, epochs=1, warmup_steps=100)
自 v4.0 版本发布以来，仍然可以使用 CrossEncoder.fit()，但它将在后台初始化 CrossEncoderTrainer。建议直接使用 Trainer，因为您可以通过 CrossEncoderTrainingArguments 获得更多控制，但依赖于 CrossEncoder.fit() 的现有训练脚本仍然应该可以工作。
如果更新后的 CrossEncoder.fit() 存在问题，您还可以通过调用 CrossEncoder.old_fit() 来获得完全旧的行为，但此方法计划在未来完全弃用。
与 SentenceTransformer 训练的比较
训练 CrossEncoder 模型与训练 SentenceTransformer 模型非常相似，但存在一些关键差异
对于
CrossEncoder训练，除了score和label之外，名为scores和labels的列也将被视为“标签列”。正如您在 Loss Overview 文档中看到的那样，某些损失需要在具有这些名称之一的列中使用特定的标签/分数。在
SentenceTransformer训练中，您不能在训练/评估数据集的列中使用输入列表（例如，文本）。对于CrossEncoder训练，您可以在列中使用（可变大小的）文本列表。例如，ListNetLoss类需要这样做。
有关训练 SentenceTransformer 模型的更多详细信息，请参阅 Sentence Transformer > Training Overview 文档。