AI 资源中心

共 339 项资源

排序方式

下载量收藏数最新

厂商

全部 01.AI Alibaba Alibaba-NLP Andycurrent Anthropic BAAI Bingsu Cohere Comfy-Org DeepSeek E-MIMIC EleutherAI FacebookAI Falconsai FinLang Google Kijai MahmoudAshraf Marqo Meta Midjourney Mistral AI NeoQuasar OpenAI Perplexity AI ProsusAI Qwen RedHatAI ResembleAI Salesforce Stability AI TinyLlama TostAI Xenova amazon apple argmaxinc autogluon cardiffnlp colbert-ir coqui cross-encoder cyankiwi daekeun-ml deepseek-ai depth-anything dima806 distilbert docling-project dphn emilyalsentzer facebook google google-bert google-t5 hexgrad hmellor intfloat jinaai jonatasgrosman k2-fsa laion llava-hf lpiccinelli meta-llama microsoft mistralai mixedbread-ai neuralmind nomic-ai nvidia openai openai-community patrickjohncyh prajjwal1 pyannote rhasspy sentence-transformers speechbrain stabilityai timm trl-internal-testing unsloth usyd-community vikhyatk xAI zai-org 商汤科技字节跳动智谱AI 月之暗面百川智能百度科大讯飞稀宇科技腾讯阶跃星辰

all-MiniLM-L6-v2

sentence-similarity

sentence-transformers

all-MiniLM-L6-v2 这是一个句子变换器模型：它将句子和段落映射到384维的稠密向量空间，可用于聚类或语义搜索等任务。

5077

Qwen3-VL-2B-Instruct

image-text-to-text

Qwen

Meet Qwen3-VL — 迄今为止通义千问系列中最强大的视觉语言模型。

686

bert-base-uncased

fill-mask

google-bert

基于掩码语言建模（MLM）目标在英语上预训练的模型。该模型首次发表于此论文，并首次在此仓库中发布。此模型不区分大小写：即对"english"和"English"不作区分。

2929

electra-base-discriminator

google

ELECTRA：以判别器而非生成器方式预训练文本编码器

389

paraphrase-multilingual-MiniLM-L12-v2

sentence-similarity

sentence-transformers

sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2

1521

ms-marco-MiniLM-L6-v2

text-ranking

cross-encoder

该模型在MS Marco段落排序任务上进行了训练。

501

bge-small-en-v1.5

feature-extraction

BAAI

733

all-mpnet-base-v2

sentence-similarity

sentence-transformers

all-mpnet-base-v2 这是一个句子变换器模型：它将句子和段落映射到768维的稠密向量空间，可用于聚类或语义搜索等任务。

1593

clip-vit-large-patch14

zero-shot-image-classification

openai

Disclaimer: The model card is taken and modified from the official CLIP repository, it can be found here.

2304

bge-m3

sentence-similarity

BAAI

更多详情请参考我们的GitHub仓库：https://github.com/FlagOpen/FlagEmbedding

3282

clip-vit-base-patch32

zero-shot-image-classification

openai

Disclaimer: The model card is taken and modified from the official CLIP repository, it can be found here.

1252

clap-htsat-fused

audio-classification

laion

Model card for CLAP: Contrastive Language-Audio Pretraining

386

roberta-large

fill-mask

FacebookAI

--- language: en tags: - exbert license: mit datasets: - bookcorpus - wikipedia --- RoBERTa 大型模型基于掩码语言建模（MLM）目标在英语上预训练的模型。该模型首次在此论文中提出，并在此仓库中首次发布。此模型区分大小写。

565

xlm-roberta-base

fill-mask

FacebookAI

XLM-RoBERTa模型在包含100种语言的2.5TB过滤CommonCrawl数据上进行了预训练。该模型由Conneau等人在论文《大规模无监督跨语言表示学习》中提出，并首次在此仓库中发布。

1128

mobilenetv3_small_100.lamb_in1k

image-classification

timm

A MobileNet-v3 image classification model. Trained on ImageNet-1k in `timm` using recipe template described below.

380

Qwen3-0.6B

text-generation

Qwen

Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groun

1547

roberta-base

fill-mask

FacebookAI

基于遮蔽语言建模（MLM）目标在英语上预训练的模型。该模型首次发表于此论文，并首次发布于此仓库。此模型区分大小写：它能识别english与English之间的差异。

892

chronos-2

time-series-forecasting

amazon

**Update Dec 30, 2025:** ☁️ Deploy Chronos-2 on Amazon SageMaker. New guide covers real-time GPU and CPU inference, serverless endpoints (run on demand, no idle costs), and batch transform for large-s

562

distilbert-base-uncased

fill-mask

distilbert

该模型是BERT基础模型的蒸馏版本。该模型首次发表于此论文中。蒸馏过程的代码可在此处获取。该模型不区分大小写：它不会区分"english"和"English"。

1169

gpt2

text-generation

openai-community

在此处测试整体生成能力：https://transformer.huggingface.co/doc/gpt2-large

3514

查看全部 242 个模型 →