From a305b032bffce0671592d897cabeca36a228f7de Mon Sep 17 00:00:00 2001 From: Ren Xuancheng Date: Fri, 29 Aug 2025 12:53:25 +0800 Subject: [PATCH] update files and links --- Chinese-CLIP-on-MUGE-Retrieval.ipynb | 2199 ++++++++++++++++++++++++++ MIT-LICENSE.txt | 2 +- README.md | 30 +- README_En.md | 26 +- cn_clip/clip/utils.py | 64 +- datasets/ImageNet-1K/label.txt | 1000 ++++++++++++ datasets/ImageNet-1K/label_cn.txt | 1000 ++++++++++++ deployment.md | 10 +- deployment_En.md | 10 +- zeroshot_dataset.md | 4 +- zeroshot_dataset_en.md | 4 +- 11 files changed, 4289 insertions(+), 60 deletions(-) create mode 100644 Chinese-CLIP-on-MUGE-Retrieval.ipynb create mode 100644 datasets/ImageNet-1K/label.txt create mode 100644 datasets/ImageNet-1K/label_cn.txt diff --git a/Chinese-CLIP-on-MUGE-Retrieval.ipynb b/Chinese-CLIP-on-MUGE-Retrieval.ipynb new file mode 100644 index 0000000..7b41b72 --- /dev/null +++ b/Chinese-CLIP-on-MUGE-Retrieval.ipynb @@ -0,0 +1,2199 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 电商多模态图文检索——Chinese-CLIP baseline全流程示例" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "ExecutionIndicator": { + "show": true + }, + "execution": { + "iopub.execute_input": "2023-01-13T07:36:25.953208Z", + "iopub.status.busy": "2023-01-13T07:36:25.952778Z", + "iopub.status.idle": "2023-01-13T07:36:25.958451Z", + "shell.execute_reply": "2023-01-13T07:36:25.957481Z", + "shell.execute_reply.started": "2023-01-13T07:36:25.953162Z" + } + }, + "source": [ + "

\n", + "
\n", + " \n", + "
\n", + "

\n", + "\n", + "Chinese-CLIP是OpenAI CLIP模型的中文版本，基于OpenAI CLIP的视觉侧参数，继续使用大规模中文原生图文数据，进行如下图所示的两阶段预训练（\\~2亿中文图文对）。旨在帮助用户快速实现中文领域的图文特征&相似度计算、跨模态检索、零样本图片分类等任务，在[电商多模态图文检索挑战赛](https://tianchi.aliyun.com/competition/entrance/532031/introduction)（即MUGE检索数据集）效果显著。我们这里提供一个基于base规模的baseline全流程实现，方便同学们更简单上手，可以在验证集取得\\~75左右的Mean Recall分数。如果大家想要深入了解和改进模型代码，提高比赛分数，欢迎在Github开源repo了解更多细节\\~\n", + "\n", + "

\n", + "
\n", + " \n", + "
\n", + "

\n", + "\n", + "**代码Github开源链接（欢迎star \\& fork🔥🔥）:** https://github.com/OFA-Sys/Chinese-CLIP \n", + " \n", + "**技术报告:** https://arxiv.org/abs/2211.01335\n", + "\n", + "**电商多模态图文检索挑战赛主页**: https://tianchi.aliyun.com/competition/entrance/532031/introduction" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 说明：\n", + "\n", + "1. 基于此notebook，可以轻松实现利用**Chinese-CLIP**图文预训练双塔表征模型，在电商多模态图文检索挑战赛完成finetune及验证集/测试集预测\n", + "2. 由于对比学习对于训练batch大小有一定的基本要求，此notebook运行至少**需要8GB显存**，请大家在GPU显存满足此要求的运行环境下下载并运行此notebook（[下载链接](https://github.com/OFA-Sys/Chinese-CLIP/blob/master/Chinese-CLIP-on-MUGE-Retrieval.ipynb)），尝试baseline流程。目前天池实验室Notebook环境暂不支持使用这个规模的显存，达摩院[ModelScope平台](https://modelscope.cn/models/damo/multi-modal_clip-vit-base-patch16_zh/summary)提供了一定的GPU机时（进入链接后右上角\"在Notebook打开\"），资源紧张的同学可以考虑利用\n", + "3. 请保证环境Pytorch版本至少在1.8.0及以上\n", + "4. Notebook中的流程较为简单，主要方便参赛选手上手跑通一个不错的base规模（视觉侧ViT-B/16，文本侧Roberta-base）baseline。如果要继续调优或改进模型，建议进一步详细参考github开源源码（尤其是[readme跨模态检索部分的文档](https://github.com/OFA-Sys/Chinese-CLIP#跨模态检索)）和Chinese-CLIP技术报告" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 准备工作\n", + "### 1. Clone代码库并准备数据目录" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-13T07:36:28.950521Z", + "iopub.status.busy": "2023-01-13T07:36:28.950171Z", + "iopub.status.idle": "2023-01-13T07:36:36.615030Z", + "shell.execute_reply": "2023-01-13T07:36:36.614150Z", + "shell.execute_reply.started": "2023-01-13T07:36:28.950487Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Cloning into 'Chinese-CLIP'...\n", + "remote: Enumerating objects: 990, done.\u001b[K\n", + "remote: Counting objects: 100% (364/364), done.\u001b[K\n", + "remote: Compressing objects: 100% (131/131), done.\u001b[K\n", + "remote: Total 990 (delta 328), reused 233 (delta 233), pack-reused 626\u001b[K\n", + "Receiving objects: 100% (990/990), 402.11 KiB | 593.00 KiB/s, done.\n", + "Resolving deltas: 100% (633/633), done.\n", + "Updating files: 100% (54/54), done.\n", + "\u001b[01;34m.\u001b[0m\n", + "├── \u001b[01;34mChinese-CLIP\u001b[0m\n", + "│   ├── \u001b[01;34massets\u001b[0m\n", + "│   │   └── \u001b[00mChinese_CLIP_logo_tp_path.svg\u001b[0m\n", + "│   ├── \u001b[01;34mcn_clip\u001b[0m\n", + "│   │   ├── \u001b[01;34mclip\u001b[0m\n", + "│   │   │   ├── \u001b[00mbert_tokenizer.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mconfiguration_bert.py\u001b[0m\n", + "│   │   │   ├── \u001b[00m__init__.py\u001b[0m\n", + "│   │   │   ├── \u001b[01;34mmodel_configs\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mRBT3-chinese.json\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mRN50.json\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mRoBERTa-wwm-ext-base-chinese.json\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mRoBERTa-wwm-ext-large-chinese.json\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mViT-B-16.json\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mViT-B-32.json\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mViT-H-14.json\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mViT-L-14-336.json\u001b[0m\n", + "│   │   │   │   └── \u001b[00mViT-L-14.json\u001b[0m\n", + "│   │   │   ├── \u001b[00mmodeling_bert.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mmodel.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mutils.py\u001b[0m\n", + "│   │   │   └── \u001b[00mvocab.txt\u001b[0m\n", + "│   │   ├── \u001b[01;34meval\u001b[0m\n", + "│   │   │   ├── \u001b[00mcvinw_zeroshot_templates.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mdata.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mevaluation.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mevaluation_tr.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mextract_features.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mimagenet_zeroshot_templates.py\u001b[0m\n", + "│   │   │   ├── \u001b[00m__init__.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mmake_topk_predictions.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mmake_topk_predictions_tr.py\u001b[0m\n", + "│   │   │   ├── \u001b[00mtransform_ir_annotation_to_tr.py\u001b[0m\n", + "│   │   │   └── \u001b[00mzeroshot_evaluation.py\u001b[0m\n", + "│   │   ├── \u001b[00m__init__.py\u001b[0m\n", + "│   │   ├── \u001b[01;34mpreprocess\u001b[0m\n", + "│   │   │   ├── \u001b[00mbuild_lmdb_dataset.py\u001b[0m\n", + "│   │   │   ├── \u001b[00m__init__.py\u001b[0m\n", + "│   │   │   └── \u001b[00mtransform_openai_pretrain_weights.py\u001b[0m\n", + "│   │   └── \u001b[01;34mtraining\u001b[0m\n", + "│   │   ├── \u001b[00mdata.py\u001b[0m\n", + "│   │   ├── \u001b[00m__init__.py\u001b[0m\n", + "│   │   ├── \u001b[00mlogger.py\u001b[0m\n", + "│   │   ├── \u001b[00mmain.py\u001b[0m\n", + "│   │   ├── \u001b[00mparams.py\u001b[0m\n", + "│   │   ├── \u001b[00mscheduler.py\u001b[0m\n", + "│   │   └── \u001b[00mtrain.py\u001b[0m\n", + "│   ├── \u001b[01;34mexamples\u001b[0m\n", + "│   │   └── \u001b[01;35mpokemon.jpeg\u001b[0m\n", + "│   ├── \u001b[00mMIT-LICENSE.txt\u001b[0m\n", + "│   ├── \u001b[00mREADME_En.md\u001b[0m\n", + "│   ├── \u001b[00mREADME.md\u001b[0m\n", + "│   ├── \u001b[00mrequirements.txt\u001b[0m\n", + "│   ├── \u001b[00mResults.md\u001b[0m\n", + "│   ├── \u001b[01;34mrun_scripts\u001b[0m\n", + "│   │   ├── \u001b[00mflickr30k_finetune_vit-b-16_rbt-base_flip.sh\u001b[0m\n", + "│   │   ├── \u001b[00mflickr30k_finetune_vit-b-16_rbt-base.sh\u001b[0m\n", + "│   │   ├── \u001b[00mmuge_finetune_vit-b-16_rbt-base_flip.sh\u001b[0m\n", + "│   │   ├── \u001b[00mmuge_finetune_vit-b-16_rbt-base.sh\u001b[0m\n", + "│   │   └── \u001b[00mzeroshot_eval.sh\u001b[0m\n", + "│   ├── \u001b[00msetup.py\u001b[0m\n", + "│   ├── \u001b[00mzeroshot_dataset_en.md\u001b[0m\n", + "│   └── \u001b[00mzeroshot_dataset.md\u001b[0m\n", + "├── \u001b[00mChinese-CLIP-on-Retrieval.ipynb\u001b[0m\n", + "└── \u001b[01;34mdatapath\u001b[0m\n", + " ├── \u001b[01;34mdatasets\u001b[0m\n", + " ├── \u001b[01;34mexperiments\u001b[0m\n", + " └── \u001b[01;34mpretrained_weights\u001b[0m\n", + "\n", + "14 directories, 54 files\n" + ] + } + ], + "source": [ + "# clone Chinese-CLIP代码库到用户目录(如果因网络问题卡在git clone这步，请interrupt该步并重试几次)：\n", + "!git clone https://github.com/OFA-Sys/Chinese-CLIP.git\n", + "\n", + "# 创建一个datapath文件夹，用于存放数据集和Chinese-CLIP参数\n", + "!mkdir -p datapath\n", + "!mkdir -p datapath/pretrained_weights # 存放预训练参数\n", + "!mkdir -p datapath/datasets # 存放数据集\n", + "!mkdir -p datapath/experiments # 存放finetune参数和日志\n", + "\n", + "!tree ." + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### 2. 安装Chinese-CLIP相关依赖" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-13T07:38:24.533603Z", + "iopub.status.busy": "2023-01-13T07:38:24.533243Z", + "iopub.status.idle": "2023-01-13T07:38:26.953741Z", + "shell.execute_reply": "2023-01-13T07:38:26.952463Z", + "shell.execute_reply.started": "2023-01-13T07:38:24.533567Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Looking in indexes: https://mirrors.aliyun.com/pypi/simple/\n", + "Requirement already satisfied: numpy in /home/pai/lib/python3.6/site-packages (from -r Chinese-CLIP/requirements.txt (line 1)) (1.19.5)\n", + "Requirement already satisfied: tqdm in /home/pai/lib/python3.6/site-packages (from -r Chinese-CLIP/requirements.txt (line 2)) (4.64.0)\n", + "Requirement already satisfied: six in /home/pai/lib/python3.6/site-packages (from -r Chinese-CLIP/requirements.txt (line 3)) (1.16.0)\n", + "Requirement already satisfied: timm in /home/pai/lib/python3.6/site-packages (from -r Chinese-CLIP/requirements.txt (line 4)) (0.6.12)\n", + "Requirement already satisfied: lmdb==1.3.0 in /home/pai/lib/python3.6/site-packages (from -r Chinese-CLIP/requirements.txt (line 5)) (1.3.0)\n", + "Requirement already satisfied: torch>=1.7.1 in /home/pai/lib/python3.6/site-packages (from -r Chinese-CLIP/requirements.txt (line 6)) (1.10.2+cu113)\n", + "Requirement already satisfied: torchvision in /home/pai/lib/python3.6/site-packages (from -r Chinese-CLIP/requirements.txt (line 7)) (0.11.3+cu113)\n", + "Requirement already satisfied: importlib-resources in /home/pai/lib/python3.6/site-packages (from tqdm->-r Chinese-CLIP/requirements.txt (line 2)) (5.4.0)\n", + "Requirement already satisfied: huggingface-hub in /home/pai/lib/python3.6/site-packages (from timm->-r Chinese-CLIP/requirements.txt (line 4)) (0.4.0)\n", + "Requirement already satisfied: pyyaml in /home/pai/lib/python3.6/site-packages (from timm->-r Chinese-CLIP/requirements.txt (line 4)) (5.4.1)\n", + "Requirement already satisfied: dataclasses in /home/pai/lib/python3.6/site-packages (from torch>=1.7.1->-r Chinese-CLIP/requirements.txt (line 6)) (0.8)\n", + "Requirement already satisfied: typing-extensions in /home/pai/lib/python3.6/site-packages (from torch>=1.7.1->-r Chinese-CLIP/requirements.txt (line 6)) (3.7.4.3)\n", + "Requirement already satisfied: pillow!=8.3.0,>=5.3.0 in /home/pai/lib/python3.6/site-packages (from torchvision->-r Chinese-CLIP/requirements.txt (line 7)) (8.4.0)\n", + "Requirement already satisfied: requests in /home/pai/lib/python3.6/site-packages (from huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (2.18.4)\n", + "Requirement already satisfied: importlib-metadata in /home/pai/lib/python3.6/site-packages (from huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (4.8.1)\n", + "Requirement already satisfied: packaging>=20.9 in /home/pai/lib/python3.6/site-packages (from huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (21.3)\n", + "Requirement already satisfied: filelock in /home/pai/lib/python3.6/site-packages (from huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (3.4.1)\n", + "Requirement already satisfied: zipp>=3.1.0 in /home/pai/lib/python3.6/site-packages (from importlib-resources->tqdm->-r Chinese-CLIP/requirements.txt (line 2)) (3.6.0)\n", + "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /home/pai/lib/python3.6/site-packages (from packaging>=20.9->huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (3.0.9)\n", + "Requirement already satisfied: chardet<3.1.0,>=3.0.2 in /home/pai/lib/python3.6/site-packages (from requests->huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (3.0.4)\n", + "Requirement already satisfied: idna<2.7,>=2.5 in /home/pai/lib/python3.6/site-packages (from requests->huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (2.6)\n", + "Requirement already satisfied: urllib3<1.23,>=1.21.1 in /home/pai/lib/python3.6/site-packages (from requests->huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (1.22)\n", + "Requirement already satisfied: certifi>=2017.4.17 in /home/pai/lib/python3.6/site-packages (from requests->huggingface-hub->timm->-r Chinese-CLIP/requirements.txt (line 4)) (2021.5.30)\n", + "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\n" + ] + } + ], + "source": [ + "# 安装Chinese-CLIP依赖包\n", + "!pip install -r Chinese-CLIP/requirements.txt" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-13T07:38:30.261373Z", + "iopub.status.busy": "2023-01-13T07:38:30.261084Z", + "iopub.status.idle": "2023-01-13T07:38:31.534984Z", + "shell.execute_reply": "2023-01-13T07:38:31.533790Z", + "shell.execute_reply.started": "2023-01-13T07:38:30.261345Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Package Version\n", + "---------------------------- ------------\n", + "absl-py 0.15.0\n", + "aiohttp 3.8.1\n", + "aiosignal 1.2.0\n", + "alibabacloud-credentials 0.2.0\n", + "alibabacloud-endpoint-util 0.0.3\n", + "alibabacloud-gateway-spi 0.0.1\n", + "alibabacloud-openapi-util 0.1.6\n", + "alibabacloud-pai-dlc20201203 1.0.0\n", + "alibabacloud-tea 0.2.9\n", + "alibabacloud-tea-openapi 0.3.3\n", + "alibabacloud-tea-util 0.3.5\n", + "alibabacloud-tea-xml 0.0.2\n", + "alipai 0.1.7\n", + "aliyun-log-python-sdk 0.7.9\n", + "aliyun-python-sdk-core 2.13.36\n", + "aliyun-python-sdk-kms 2.15.0\n", + "aliyun-python-sdk-sts 3.1.0\n", + "apache-beam 2.15.0\n", + "apache-flink 1.10.1\n", + "argcomplete 2.0.0\n", + "argon2-cffi 21.3.0\n", + "argon2-cffi-bindings 21.2.0\n", + "asn1crypto 0.24.0\n", + "astor 0.8.1\n", + "astroid 2.11.5\n", + "astunparse 1.6.3\n", + "async-generator 1.10\n", + "async-timeout 4.0.2\n", + "asynctest 0.13.0\n", + "attrs 21.4.0\n", + "autopep8 1.6.0\n", + "avro-python3 1.9.1\n", + "backcall 0.2.0\n", + "backports.zoneinfo 0.2.1\n", + "bleach 4.1.0\n", + "brotlipy 0.7.0\n", + "cached-property 1.5.2\n", + "cachetools 4.2.4\n", + "certifi 2021.5.30\n", + "cffi 1.15.0\n", + "chardet 3.0.4\n", + "charset-normalizer 2.0.12\n", + "clang 5.0\n", + "cloudpickle 1.2.2\n", + "colorama 0.4.4\n", + "common-io 0.3.0\n", + "conda 4.10.3\n", + "conda-package-handling 1.7.3\n", + "configparser 5.2.0\n", + "contextlib2 21.6.0\n", + "crcmod 1.7\n", + "cryptography 37.0.2\n", + "cvxopt 1.3.0\n", + "cycler 0.11.0\n", + "Cython 0.29.24\n", + "dataclasses 0.8\n", + "dateparser 1.1.1\n", + "decorator 4.4.2\n", + "defusedxml 0.7.1\n", + "deprecation 2.1.0\n", + "dill 0.2.9\n", + "docopt 0.6.2\n", + "eas-prediction 0.13\n", + "easy-rec 0.1.6\n", + "elastic-transport 8.1.2\n", + "elasticsearch 8.2.0\n", + "entrypoints 0.4\n", + "fastavro 0.21.24\n", + "filelock 3.4.1\n", + "flake8 4.0.1\n", + "flatbuffers 1.12\n", + "frozenlist 1.2.0\n", + "future 0.18.2\n", + "gast 0.4.0\n", + "google-auth 1.35.0\n", + "google-auth-oauthlib 0.4.6\n", + "google-pasta 0.2.0\n", + "graphviz 0.19.1\n", + "grpcio 1.46.3\n", + "h5py 3.1.0\n", + "hdfs 2.7.0\n", + "htmlmin 0.1.12\n", + "httplib2 0.12.0\n", + "huggingface-hub 0.4.0\n", + "hyperopt 0.1.2\n", + "idna 2.6\n", + "idna-ssl 1.1.0\n", + "ImageHash 4.2.1\n", + "importlib-metadata 4.8.1\n", + "importlib-resources 5.4.0\n", + "ipykernel 5.5.6\n", + "ipython 7.16.3\n", + "ipython-genutils 0.2.0\n", + "ipywidgets 7.7.0\n", + "isort 5.10.1\n", + "jedi 0.17.2\n", + "Jinja2 3.0.3\n", + "jmespath 0.10.0\n", + "joblib 1.1.0\n", + "json-tricks 3.15.5\n", + "jsonschema 4.0.0\n", + "jupyter-client 7.1.2\n", + "jupyter-core 4.9.2\n", + "jupyterlab-pygments 0.1.2\n", + "jupyterlab-widgets 1.1.0\n", + "keras 2.6.0\n", + "Keras-Preprocessing 1.1.2\n", + "kiwisolver 1.3.1\n", + "lazy-object-proxy 1.7.1\n", + "llvmlite 0.36.0\n", + "lmdb 1.3.0\n", + "mamba 0.15.3\n", + "Markdown 3.3.7\n", + "MarkupSafe 2.0.1\n", + "matplotlib 3.3.4\n", + "mccabe 0.6.1\n", + "minio 7.1.8\n", + "missingno 0.4.2\n", + "mistune 0.8.4\n", + "mock 2.0.0\n", + "multidict 5.2.0\n", + "multimethod 1.4\n", + "nbclient 0.5.9\n", + "nbconvert 6.0.7\n", + "nbformat 5.1.3\n", + "nest-asyncio 1.5.5\n", + "networkx 2.7.1\n", + "notebook 6.4.10\n", + "numba 0.53.1\n", + "numpy 1.19.5\n", + "oauth2client 3.0.0\n", + "oauthlib 3.2.0\n", + "olefile 0.46\n", + "opencv-contrib-python 4.5.5.64\n", + "opencv-python 4.5.5.64\n", + "opt-einsum 3.3.0\n", + "oss2 2.15.0\n", + "packaging 21.3\n", + "pai-nni 2.6\n", + "pandas 1.1.5\n", + "pandas-profiling 3.1.0\n", + "pandocfilters 1.5.0\n", + "parso 0.7.1\n", + "patsy 0.5.2\n", + "pbr 5.9.0\n", + "pexpect 4.8.0\n", + "phik 0.11.2\n", + "pickleshare 0.7.5\n", + "Pillow 8.4.0\n", + "pip 21.3.1\n", + "platformdirs 2.4.0\n", + "plotly 5.8.0\n", + "prettytable 2.5.0\n", + "prometheus-client 0.14.1\n", + "prompt-toolkit 3.0.29\n", + "protobuf 3.19.4\n", + "psutil 5.9.1\n", + "ptyprocess 0.7.0\n", + "py4j 0.10.8.1\n", + "pyalink-public 1.5.1\n", + "pyarrow 0.13.0\n", + "pyasn1 0.4.8\n", + "pyasn1-modules 0.2.8\n", + "pybind11 2.7.0\n", + "pycodestyle 2.8.0\n", + "pycosat 0.6.3\n", + "pycparser 2.21\n", + "pycryptodome 3.14.1\n", + "pydantic 1.8.2\n", + "pydot 1.4.2\n", + "pyflakes 2.4.0\n", + "Pygments 2.12.0\n", + "pylint 2.13.9\n", + "pymongo 3.12.3\n", + "pyodps 0.11.0\n", + "pyOpenSSL 18.0.0\n", + "pyparsing 3.0.9\n", + "pyrsistent 0.18.0\n", + "PySocks 1.6.8\n", + "python-dateutil 2.8.2\n", + "PythonWebHDFS 0.2.3\n", + "pytz 2022.1\n", + "pytz-deprecation-shim 0.1.0.post0\n", + "PyWavelets 1.1.1\n", + "PyYAML 5.4.1\n", + "pyzmq 23.0.0\n", + "regex 2022.3.2\n", + "requests 2.18.4\n", + "requests-oauthlib 1.3.1\n", + "responses 0.17.0\n", + "rsa 4.8\n", + "ruamel_yaml 0.15.37\n", + "ruamel-yaml-conda 0.15.100\n", + "Rx 3.2.0\n", + "schema 0.7.5\n", + "scikit-learn 0.24.2\n", + "scipy 1.5.4\n", + "seaborn 0.11.2\n", + "Send2Trash 1.8.0\n", + "setuptools 58.0.4\n", + "simplejson 3.17.6\n", + "six 1.16.0\n", + "statsmodels 0.12.2\n", + "tangled-up-in-unicode 0.1.0\n", + "tenacity 8.0.1\n", + "tensorboard 2.6.0\n", + "tensorboard-data-server 0.6.1\n", + "tensorboard-plugin-wit 1.8.1\n", + "tensorflow 2.6.2\n", + "tensorflow-estimator 2.6.0\n", + "tensorflow-gpu 2.6.0\n", + "tensorflow-io 0.21.0\n", + "tensorflow-io-gcs-filesystem 0.21.0\n", + "termcolor 1.1.0\n", + "terminado 0.13.0\n", + "testpath 0.6.0\n", + "threadpoolctl 3.1.0\n", + "timm 0.6.12\n", + "toml 0.10.2\n", + "tomli 1.2.3\n", + "torch 1.10.2+cu113\n", + "torchaudio 0.10.2+cu113\n", + "torchvision 0.11.3+cu113\n", + "tornado 6.1\n", + "tqdm 4.64.0\n", + "traitlets 4.3.3\n", + "typed-ast 1.5.4\n", + "typing-extensions 3.7.4.3\n", + "tzdata 2022.1\n", + "tzlocal 4.2\n", + "urllib3 1.22\n", + "visions 0.7.4\n", + "wcwidth 0.2.5\n", + "webencodings 0.5.1\n", + "websockets 9.1\n", + "Werkzeug 2.0.3\n", + "wheel 0.37.1\n", + "widgetsnbextension 3.6.0\n", + "wrapt 1.12.1\n", + "xgboost 1.5.2\n", + "xlrd 2.0.1\n", + "xmltodict 0.13.0\n", + "yarl 1.7.2\n", + "yq 2.13.0\n", + "zipp 3.6.0\n" + ] + } + ], + "source": [ + "# 查看当前kernel下已安装的包 list packages\n", + "!pip list --format=columns" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### 3. 下载Chinese-CLIP预训练参数" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": { + "ExecutionIndicator": { + "show": false + }, + "execution": { + "iopub.execute_input": "2023-01-13T07:38:37.732489Z", + "iopub.status.busy": "2023-01-13T07:38:37.732172Z", + "iopub.status.idle": "2023-01-13T07:39:18.308756Z", + "shell.execute_reply": "2023-01-13T07:39:18.307760Z", + "shell.execute_reply.started": "2023-01-13T07:38:37.732449Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [] + } + ], + "source": [ + "# 下载base规模Chinese-CLIP预训练参数：\n", + "!wget https://www.modelscope.cn/models/AI-ModelScope/chinese-clip-vit-base-patch16/resolve/master/clip_cn_vit-b-16.pt\n", + "!mv clip_cn_vit-b-16.pt datapath/pretrained_weights/" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### 4. 预处理数据" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "ExecutionIndicator": { + "show": false + }, + "execution": { + "iopub.execute_input": "2023-01-13T07:41:28.580949Z", + "iopub.status.busy": "2023-01-13T07:41:28.580661Z", + "iopub.status.idle": "2023-01-13T07:43:34.116492Z", + "shell.execute_reply": "2023-01-13T07:43:34.115666Z", + "shell.execute_reply.started": "2023-01-13T07:41:28.580909Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "--2023-01-13 15:41:28-- https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/resolve/main/MUGE.zip\n", + "...\n", + "HTTP request sent, awaiting response... 200 OK\n", + "Length: 2166554702 (2.0G) [application/zip]\n", + "Saving to: ‘MUGE.zip’\n", + "\n", + "MUGE.zip 100%[===================>] 2.02G 26.7MB/s in 89s \n", + "\n", + "2023-01-13 15:42:58 (23.1 MB/s) - ‘MUGE.zip’ saved [2166554702/2166554702]\n", + "\n", + "/mnt/workspace/tianchi/datapath/datasets\n", + "Archive: MUGE.zip\n", + " creating: MUGE/\n", + " creating: MUGE/lmdb/\n", + " creating: MUGE/lmdb/train/\n", + " creating: MUGE/lmdb/train/imgs/\n", + " inflating: MUGE/lmdb/train/imgs/data.mdb \n", + " inflating: MUGE/lmdb/train/imgs/lock.mdb \n", + " creating: MUGE/lmdb/train/pairs/\n", + " inflating: MUGE/lmdb/train/pairs/data.mdb \n", + " inflating: MUGE/lmdb/train/pairs/lock.mdb \n", + " creating: MUGE/lmdb/valid/\n", + " creating: MUGE/lmdb/valid/imgs/\n", + " inflating: MUGE/lmdb/valid/imgs/data.mdb \n", + " inflating: MUGE/lmdb/valid/imgs/lock.mdb \n", + " creating: MUGE/lmdb/valid/pairs/\n", + " inflating: MUGE/lmdb/valid/pairs/data.mdb \n", + " inflating: MUGE/lmdb/valid/pairs/lock.mdb \n", + " inflating: MUGE/train_imgs.tsv \n", + " inflating: MUGE/train_texts.jsonl \n", + " inflating: MUGE/valid_imgs.tsv \n", + " inflating: MUGE/valid_texts.jsonl \n", + "/mnt/workspace/tianchi\n", + "\u001b[01;34mdatapath/\u001b[0m\n", + "├── \u001b[01;34mdatasets\u001b[0m\n", + "│   ├── \u001b[01;34mMUGE\u001b[0m\n", + "│   │   ├── \u001b[01;34mlmdb\u001b[0m\n", + "│   │   │   ├── \u001b[01;34mtrain\u001b[0m\n", + "│   │   │   │   ├── \u001b[01;34mimgs\u001b[0m\n", + "│   │   │   │   │   ├── \u001b[00mdata.mdb\u001b[0m\n", + "│   │   │   │   │   └── \u001b[00mlock.mdb\u001b[0m\n", + "│   │   │   │   └── \u001b[01;34mpairs\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mdata.mdb\u001b[0m\n", + "│   │   │   │   └── \u001b[00mlock.mdb\u001b[0m\n", + "│   │   │   └── \u001b[01;34mvalid\u001b[0m\n", + "│   │   │   ├── \u001b[01;34mimgs\u001b[0m\n", + "│   │   │   │   ├── \u001b[00mdata.mdb\u001b[0m\n", + "│   │   │   │   └── \u001b[00mlock.mdb\u001b[0m\n", + "│   │   │   └── \u001b[01;34mpairs\u001b[0m\n", + "│   │   │   ├── \u001b[00mdata.mdb\u001b[0m\n", + "│   │   │   └── \u001b[00mlock.mdb\u001b[0m\n", + "│   │   ├── \u001b[00mtrain_imgs.tsv\u001b[0m\n", + "│   │   ├── \u001b[00mtrain_texts.jsonl\u001b[0m\n", + "│   │   ├── \u001b[00mvalid_imgs.tsv\u001b[0m\n", + "│   │   └── \u001b[00mvalid_texts.jsonl\u001b[0m\n", + "│   └── \u001b[01;31mMUGE.zip\u001b[0m\n", + "├── \u001b[01;34mexperiments\u001b[0m\n", + "└── \u001b[01;34mpretrained_weights\u001b[0m\n", + " └── \u001b[00mclip_cn_vit-b-16.pt\u001b[0m\n", + "\n", + "11 directories, 14 files\n" + ] + } + ], + "source": [ + "# 详细的预处理数据流程，请参见github readme: https://github.com/OFA-Sys/Chinese-CLIP#数据集格式预处理\n", + "# 我们这里方便起见，直接下载已经预处理好（处理成LMDB格式）的电商检索数据集\n", + "!wget https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/resolve/main/MUGE.zip\n", + "!mv MUGE.zip datapath/datasets/\n", + "%cd datapath/datasets/\n", + "!unzip MUGE.zip\n", + "%cd ../..\n", + "!tree datapath/" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-13T07:43:39.365603Z", + "iopub.status.busy": "2023-01-13T07:43:39.365196Z", + "iopub.status.idle": "2023-01-13T07:43:39.911726Z", + "shell.execute_reply": "2023-01-13T07:43:39.911099Z", + "shell.execute_reply.started": "2023-01-13T07:43:39.365556Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "{\"text_id\": 248816, \"text\": \"圣诞 抱枕\", \"image_ids\": [1006938, 561749, 936929, 286314, 141999, 183846]}\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# 查看数据样例\n", + "!head -n 1 datapath/datasets/MUGE/valid_texts.jsonl # 验证集第一条样本\n", + "\n", + "import lmdb\n", + "import base64\n", + "from io import BytesIO\n", + "from PIL import Image\n", + "\n", + "image_ids = [286314, 141999, 183846]\n", + "\n", + "lmdb_imgs = \"datapath/datasets/MUGE/lmdb/valid/imgs\"\n", + "env_imgs = lmdb.open(lmdb_imgs, readonly=True, create=False, lock=False, readahead=False, meminit=False)\n", + "txn_imgs = env_imgs.begin(buffers=True)\n", + "for image_id in image_ids:\n", + " image_b64 = txn_imgs.get(\"{}\".format(image_id).encode('utf-8')).tobytes()\n", + " img = Image.open(BytesIO(base64.urlsafe_b64decode(image_b64)))\n", + " img.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 运行finetune训练" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-13T07:43:47.598366Z", + "iopub.status.busy": "2023-01-13T07:43:47.598084Z", + "iopub.status.idle": "2023-01-13T07:43:47.603552Z", + "shell.execute_reply": "2023-01-13T07:43:47.602812Z", + "shell.execute_reply.started": "2023-01-13T07:43:47.598340Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "/mnt/workspace/tianchi/Chinese-CLIP\n" + ] + } + ], + "source": [ + "# 进入代码工作区目录\n", + "%cd Chinese-CLIP/" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "下面我们开始运行finetune，说明如下：\n", + "+ 下面的代码改编自代码库中的shell运行脚本: Chinese-CLIP/run_scripts/muge_finetune_vit-b-16_rbt-base.sh\n", + "+ **详细的finetune各配置项，请参见: https://github.com/OFA-Sys/Chinese-CLIP#模型finetune**\n", + "+ 执行训练的python代码，请参见: Chinese-CLIP/cn_clip/training/main.py\n", + "+ 模型实现的python代码，请参见: Chinese-CLIP/cn_clip/clip/model.py 以及比赛官方学习视频对于模型实现的介绍\n", + "+ 按照如下的超参进行finetune，要求机器需要至少达到8G显存，finetune时间大约需要40min左右（V100上测试）\n", + "+ **请注意finetune中间给出的验证集准确率，是验证集batch内部，图文对之间的Recall@1正确率，与比赛指标要求的从整个验证集/测试集图片池召回的Recall@1并不是同一个指标。这里仅用于观察训练趋势，如果要评测模型效果，请完整运行下文的特征提取、KNN召回和计算Recall流程**\n", + "+ 对比学习的训练收敛和稳定性和总batch size相关。如您使用较小的batch size（比如我们这里的48 per-GPU * 1 GPU），建议使用较小的学习率（如我们这里的3e-6）。我们推荐使用更多的GPU和更大的batch size以取得更好的效果。\n", + "+ 训练完成后如果输出打印EOFError，无视即可" + ] + }, + { + "cell_type": "code", + "execution_count": 37, + "metadata": { + "ExecutionIndicator": { + "show": false + }, + "execution": { + "iopub.execute_input": "2023-01-14T12:31:56.709738Z", + "iopub.status.busy": "2023-01-14T12:31:56.709466Z", + "iopub.status.idle": "2023-01-14T12:31:56.720085Z", + "shell.execute_reply": "2023-01-14T12:31:56.719253Z", + "shell.execute_reply.started": "2023-01-14T12:31:56.709713Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;python3 -m torch.distributed.launch --nproc_per_node=1 --nnodes=1 --node_rank=0 --master_addr=localhost --master_port=8514 cn_clip/training/main.py --train-data=../datapath//datasets/MUGE/lmdb/train --val-data=../datapath//datasets/MUGE/lmdb/valid --num-workers=4 --valid-num-workers=4 --resume=../datapath//pretrained_weights/clip_cn_vit-b-16.pt --reset-data-offset --reset-optimizer --logs=../datapath//experiments/ --name=muge_finetune_vit-b-16_roberta-base_bs48_1gpu --save-step-frequency=999999 --save-epoch-frequency=1 --log-interval=10 --report-training-batch-acc --context-length=52 --warmup=100 --batch-size=48 --valid-batch-size=48 --valid-step-interval=1000 --valid-epoch-interval=1 --lr=3e-06 --wd=0.001 --max-epochs=1 --vision-model=ViT-B-16 --use-augment --grad-checkpointing --text-model=RoBERTa-wwm-ext-base-chinese\n", + "\n" + ] + } + ], + "source": [ + "# 准备finetune相关配置，详见https://github.com/OFA-Sys/Chinese-CLIP#模型finetune\n", + "# 指定机器数 & 卡数\n", + "GPUS_PER_NODE=1 # 卡数\n", + "WORKER_CNT=1 # 机器数\n", + "MASTER_ADDR=\"localhost\"\n", + "MASTER_PORT=8514 # 同台机器同时起多个任务，请分别分配不同的端口号\n", + "RANK=0 \n", + "\n", + "# 刚刚创建过的目录，存放了预训练参数和预处理好的数据集\n", + "DATAPATH=\"../datapath/\"\n", + "\n", + "# 指定LMDB格式的训练集和验证集路径（存放了LMDB格式的图片和图文对数据）\n", + "train_data=f\"{DATAPATH}/datasets/MUGE/lmdb/train\"\n", + "val_data=f\"{DATAPATH}/datasets/MUGE/lmdb/valid\"\n", + "num_workers=4 # 训练集pytorch dataloader的进程数，设置为>0，以减小训练时读取数据的时间开销\n", + "valid_num_workers=4 # 验证集pytorch dataloader的进程数，设置为>0，以减小验证时读取数据的时间开销\n", + "\n", + "# 指定刚刚下载好的Chinese-CLIP预训练权重的路径\n", + "resume=f\"{DATAPATH}/pretrained_weights/clip_cn_vit-b-16.pt\"\n", + "reset_data_offset=\"--reset-data-offset\" # 从头读取训练数据\n", + "reset_optimizer=\"--reset-optimizer\" # 重新初始化AdamW优化器\n", + "\n", + "# 指定输出相关配置\n", + "output_base_dir=f\"{DATAPATH}/experiments/\"\n", + "name=\"muge_finetune_vit-b-16_roberta-base_bs48_1gpu\" # finetune超参、日志、ckpt将保存在../datapath/experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/\n", + "save_step_frequency=999999 # disable it\n", + "save_epoch_frequency=1 # 每轮保存一个finetune ckpt\n", + "log_interval=10 # 日志打印间隔步数\n", + "report_training_batch_acc=\"--report-training-batch-acc\" # 训练中，报告训练batch的in-batch准确率\n", + "\n", + "# 指定训练超参数\n", + "context_length=52 # 序列长度，这里指定为Chinese-CLIP默认的52\n", + "warmup=100 # warmup步数\n", + "batch_size=48 # 训练单卡batch size\n", + "valid_batch_size=48 # 验证单卡batch size\n", + "lr=3e-6 # 学习率，因为这里我们使用的对比学习batch size很小，所以对应的学习率也调低一些\n", + "wd=0.001 # weight decay\n", + "max_epochs=1 # 训练轮数，也可通过--max-steps指定训练步数\n", + "valid_step_interval=1000 # 验证步数间隔\n", + "valid_epoch_interval=1 # 验证轮数间隔\n", + "vision_model=\"ViT-B-16\" # 指定视觉侧结构为ViT-B/16\n", + "text_model=\"RoBERTa-wwm-ext-base-chinese\" # 指定文本侧结构为RoBERTa-base\n", + "use_augment=\"--use-augment\" # 对图像使用数据增强\n", + "grad_checkpointing=\"--grad-checkpointing\" # 激活重计算策略，用更多训练时间换取更小的显存开销\n", + "\n", + "run_command = \"export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\" + \\\n", + "f\"\"\"\n", + "python3 -m torch.distributed.launch --nproc_per_node={GPUS_PER_NODE} --nnodes={WORKER_CNT} --node_rank={RANK} \\\n", + " --master_addr={MASTER_ADDR} --master_port={MASTER_PORT} cn_clip/training/main.py \\\n", + " --train-data={train_data} \\\n", + " --val-data={val_data} \\\n", + " --num-workers={num_workers} \\\n", + " --valid-num-workers={valid_num_workers} \\\n", + " --resume={resume} \\\n", + " {reset_data_offset} \\\n", + " {reset_optimizer} \\\n", + " --logs={output_base_dir} \\\n", + " --name={name} \\\n", + " --save-step-frequency={save_step_frequency} \\\n", + " --save-epoch-frequency={save_epoch_frequency} \\\n", + " --log-interval={log_interval} \\\n", + " {report_training_batch_acc} \\\n", + " --context-length={context_length} \\\n", + " --warmup={warmup} \\\n", + " --batch-size={batch_size} \\\n", + " --valid-batch-size={valid_batch_size} \\\n", + " --valid-step-interval={valid_step_interval} \\\n", + " --valid-epoch-interval={valid_epoch_interval} \\\n", + " --lr={lr} \\\n", + " --wd={wd} \\\n", + " --max-epochs={max_epochs} \\\n", + " --vision-model={vision_model} \\\n", + " {use_augment} \\\n", + " {grad_checkpointing} \\\n", + " --text-model={text_model}\n", + "\"\"\".lstrip()\n", + "print(run_command)" + ] + }, + { + "cell_type": "code", + "execution_count": 38, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-14T12:32:00.252547Z", + "iopub.status.busy": "2023-01-14T12:32:00.252205Z", + "iopub.status.idle": "2023-01-14T13:18:44.855512Z", + "shell.execute_reply": "2023-01-14T13:18:44.854585Z", + "shell.execute_reply.started": "2023-01-14T12:32:00.252513Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "/home/pai/lib/python3.6/site-packages/torch/distributed/launch.py:186: FutureWarning: The module torch.distributed.launch is deprecated\n", + "and will be removed in future. Use torchrun.\n", + "Note that --use_env is set by default in torchrun.\n", + "If your script expects `--local_rank` argument to be set, please\n", + "change it to read from `os.environ['LOCAL_RANK']` instead. See \n", + "https://pytorch.org/docs/stable/distributed.html#launch-utility for \n", + "further instructions\n", + "\n", + " FutureWarning,\n", + "/home/pai/lib/python3.6/site-packages/OpenSSL/crypto.py:12: CryptographyDeprecationWarning: Python 3.6 is no longer supported by the Python core team. Therefore, support for it is deprecated in cryptography and will be removed in a future release.\n", + " from cryptography import x509\n", + "Loading vision model config from cn_clip/clip/model_configs/ViT-B-16.json\n", + "Loading text model config from cn_clip/clip/model_configs/RoBERTa-wwm-ext-base-chinese.json\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | Grad-checkpointing activated.\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | train LMDB file contains 129380 images and 250314 pairs.\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | val LMDB file contains 29806 images and 30588 pairs.\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | Params:\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | aggregate: True\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | batch_size: 48\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | bert_weight_path: None\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | beta1: 0.9\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | beta2: 0.98\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | checkpoint_path: ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | clip_weight_path: None\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | context_length: 52\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | debug: False\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | device: cuda:0\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | eps: 1e-06\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | freeze_vision: False\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | grad_checkpointing: True\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | local_device_rank: 0\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | local_rank: 0\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | log_interval: 10\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | log_level: 20\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | log_path: ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/out_2023-01-14-12-32-03.log\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | logs: ../datapath//experiments/\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | lr: 3e-06\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | mask_ratio: 0\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | max_epochs: 1\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | max_steps: 5215\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | name: muge_finetune_vit-b-16_roberta-base_bs48_1gpu\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | num_workers: 4\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | precision: amp\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | rank: 0\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | report_training_batch_acc: True\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | reset_data_offset: True\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | reset_optimizer: True\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | resume: ../datapath//pretrained_weights/clip_cn_vit-b-16.pt\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | save_epoch_frequency: 1\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | save_step_frequency: 999999\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | seed: 123\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | skip_aggregate: False\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | skip_scheduler: False\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | text_model: RoBERTa-wwm-ext-base-chinese\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | train_data: ../datapath//datasets/MUGE/lmdb/train\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | use_augment: True\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | use_bn_sync: False\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | val_data: ../datapath//datasets/MUGE/lmdb/valid\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | valid_batch_size: 48\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | valid_epoch_interval: 1\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | valid_num_workers: 4\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | valid_step_interval: 1000\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | vision_model: ViT-B-16\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | warmup: 100\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | wd: 0.001\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | world_size: 1\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | Use GPU: 0 for training\n", + "2023-01-14,20:32:11 | INFO | Rank 0 | => begin to load checkpoint '../datapath//pretrained_weights/clip_cn_vit-b-16.pt'\n", + "2023-01-14,20:32:12 | INFO | Rank 0 | => loaded checkpoint '../datapath//pretrained_weights/clip_cn_vit-b-16.pt' (epoch 15 @ 0 steps)\n", + "2023-01-14,20:32:14 | INFO | Rank 0 | Reducer buckets have been rebuilt in this iteration.\n", + "2023-01-14,20:32:17 | INFO | Rank 0 | Global Steps: 10/5215 | Train Epoch: 1 [480/250320 (0%)] | Loss: 0.551259 | Image2Text Acc: 81.25 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:21 | INFO | Rank 0 | Global Steps: 20/5215 | Train Epoch: 1 [960/250320 (0%)] | Loss: 0.655716 | Image2Text Acc: 79.17 | Text2Image Acc: 79.17 | Data Time: 0.020s | Batch Time: 0.360s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:24 | INFO | Rank 0 | Global Steps: 30/5215 | Train Epoch: 1 [1440/250320 (1%)] | Loss: 0.894897 | Image2Text Acc: 81.25 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.360s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:28 | INFO | Rank 0 | Global Steps: 40/5215 | Train Epoch: 1 [1920/250320 (1%)] | Loss: 0.891962 | Image2Text Acc: 81.25 | Text2Image Acc: 75.00 | Data Time: 0.020s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:31 | INFO | Rank 0 | Global Steps: 50/5215 | Train Epoch: 1 [2400/250320 (1%)] | Loss: 0.711004 | Image2Text Acc: 77.08 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.360s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:35 | INFO | Rank 0 | Global Steps: 60/5215 | Train Epoch: 1 [2880/250320 (1%)] | Loss: 0.602132 | Image2Text Acc: 81.25 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:39 | INFO | Rank 0 | Global Steps: 70/5215 | Train Epoch: 1 [3360/250320 (1%)] | Loss: 0.429699 | Image2Text Acc: 85.42 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:42 | INFO | Rank 0 | Global Steps: 80/5215 | Train Epoch: 1 [3840/250320 (2%)] | Loss: 0.602740 | Image2Text Acc: 81.25 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:46 | INFO | Rank 0 | Global Steps: 90/5215 | Train Epoch: 1 [4320/250320 (2%)] | Loss: 0.612829 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.024s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:50 | INFO | Rank 0 | Global Steps: 100/5215 | Train Epoch: 1 [4800/250320 (2%)] | Loss: 0.393124 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:53 | INFO | Rank 0 | Global Steps: 110/5215 | Train Epoch: 1 [5280/250320 (2%)] | Loss: 0.642643 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.018s | Batch Time: 0.360s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:32:57 | INFO | Rank 0 | Global Steps: 120/5215 | Train Epoch: 1 [5760/250320 (2%)] | Loss: 0.751711 | Image2Text Acc: 79.17 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:01 | INFO | Rank 0 | Global Steps: 130/5215 | Train Epoch: 1 [6240/250320 (2%)] | Loss: 0.654346 | Image2Text Acc: 85.42 | Text2Image Acc: 81.25 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:04 | INFO | Rank 0 | Global Steps: 140/5215 | Train Epoch: 1 [6720/250320 (3%)] | Loss: 0.744869 | Image2Text Acc: 77.08 | Text2Image Acc: 81.25 | Data Time: 0.023s | Batch Time: 0.379s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:08 | INFO | Rank 0 | Global Steps: 150/5215 | Train Epoch: 1 [7200/250320 (3%)] | Loss: 0.735694 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.370s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:12 | INFO | Rank 0 | Global Steps: 160/5215 | Train Epoch: 1 [7680/250320 (3%)] | Loss: 0.583934 | Image2Text Acc: 79.17 | Text2Image Acc: 79.17 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:15 | INFO | Rank 0 | Global Steps: 170/5215 | Train Epoch: 1 [8160/250320 (3%)] | Loss: 0.532281 | Image2Text Acc: 83.33 | Text2Image Acc: 79.17 | Data Time: 0.023s | Batch Time: 0.365s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:19 | INFO | Rank 0 | Global Steps: 180/5215 | Train Epoch: 1 [8640/250320 (3%)] | Loss: 0.340221 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:23 | INFO | Rank 0 | Global Steps: 190/5215 | Train Epoch: 1 [9120/250320 (4%)] | Loss: 0.508061 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:26 | INFO | Rank 0 | Global Steps: 200/5215 | Train Epoch: 1 [9600/250320 (4%)] | Loss: 0.459738 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:30 | INFO | Rank 0 | Global Steps: 210/5215 | Train Epoch: 1 [10080/250320 (4%)] | Loss: 0.673064 | Image2Text Acc: 83.33 | Text2Image Acc: 75.00 | Data Time: 0.021s | Batch Time: 0.379s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:34 | INFO | Rank 0 | Global Steps: 220/5215 | Train Epoch: 1 [10560/250320 (4%)] | Loss: 0.335541 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.021s | Batch Time: 0.369s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:37 | INFO | Rank 0 | Global Steps: 230/5215 | Train Epoch: 1 [11040/250320 (4%)] | Loss: 0.246524 | Image2Text Acc: 93.75 | Text2Image Acc: 97.92 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:41 | INFO | Rank 0 | Global Steps: 240/5215 | Train Epoch: 1 [11520/250320 (5%)] | Loss: 0.867559 | Image2Text Acc: 75.00 | Text2Image Acc: 77.08 | Data Time: 0.020s | Batch Time: 0.375s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:45 | INFO | Rank 0 | Global Steps: 250/5215 | Train Epoch: 1 [12000/250320 (5%)] | Loss: 0.892492 | Image2Text Acc: 77.08 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:48 | INFO | Rank 0 | Global Steps: 260/5215 | Train Epoch: 1 [12480/250320 (5%)] | Loss: 0.289513 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:52 | INFO | Rank 0 | Global Steps: 270/5215 | Train Epoch: 1 [12960/250320 (5%)] | Loss: 0.897694 | Image2Text Acc: 77.08 | Text2Image Acc: 77.08 | Data Time: 0.021s | Batch Time: 0.371s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:56 | INFO | Rank 0 | Global Steps: 280/5215 | Train Epoch: 1 [13440/250320 (5%)] | Loss: 0.518537 | Image2Text Acc: 81.25 | Text2Image Acc: 81.25 | Data Time: 0.020s | Batch Time: 0.364s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:33:59 | INFO | Rank 0 | Global Steps: 290/5215 | Train Epoch: 1 [13920/250320 (6%)] | Loss: 0.797377 | Image2Text Acc: 72.92 | Text2Image Acc: 77.08 | Data Time: 0.021s | Batch Time: 0.364s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:03 | INFO | Rank 0 | Global Steps: 300/5215 | Train Epoch: 1 [14400/250320 (6%)] | Loss: 0.523968 | Image2Text Acc: 77.08 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.368s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:07 | INFO | Rank 0 | Global Steps: 310/5215 | Train Epoch: 1 [14880/250320 (6%)] | Loss: 0.480399 | Image2Text Acc: 89.58 | Text2Image Acc: 77.08 | Data Time: 0.019s | Batch Time: 0.383s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:11 | INFO | Rank 0 | Global Steps: 320/5215 | Train Epoch: 1 [15360/250320 (6%)] | Loss: 0.580155 | Image2Text Acc: 83.33 | Text2Image Acc: 75.00 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:14 | INFO | Rank 0 | Global Steps: 330/5215 | Train Epoch: 1 [15840/250320 (6%)] | Loss: 0.424913 | Image2Text Acc: 87.50 | Text2Image Acc: 81.25 | Data Time: 0.024s | Batch Time: 0.366s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:18 | INFO | Rank 0 | Global Steps: 340/5215 | Train Epoch: 1 [16320/250320 (7%)] | Loss: 0.671518 | Image2Text Acc: 81.25 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.381s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:22 | INFO | Rank 0 | Global Steps: 350/5215 | Train Epoch: 1 [16800/250320 (7%)] | Loss: 0.533927 | Image2Text Acc: 79.17 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:25 | INFO | Rank 0 | Global Steps: 360/5215 | Train Epoch: 1 [17280/250320 (7%)] | Loss: 0.351673 | Image2Text Acc: 93.75 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:29 | INFO | Rank 0 | Global Steps: 370/5215 | Train Epoch: 1 [17760/250320 (7%)] | Loss: 0.423091 | Image2Text Acc: 87.50 | Text2Image Acc: 81.25 | Data Time: 0.021s | Batch Time: 0.370s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:33 | INFO | Rank 0 | Global Steps: 380/5215 | Train Epoch: 1 [18240/250320 (7%)] | Loss: 0.391531 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:37 | INFO | Rank 0 | Global Steps: 390/5215 | Train Epoch: 1 [18720/250320 (7%)] | Loss: 0.566264 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.024s | Batch Time: 0.380s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:40 | INFO | Rank 0 | Global Steps: 400/5215 | Train Epoch: 1 [19200/250320 (8%)] | Loss: 0.520908 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.024s | Batch Time: 0.370s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:44 | INFO | Rank 0 | Global Steps: 410/5215 | Train Epoch: 1 [19680/250320 (8%)] | Loss: 0.530346 | Image2Text Acc: 85.42 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:48 | INFO | Rank 0 | Global Steps: 420/5215 | Train Epoch: 1 [20160/250320 (8%)] | Loss: 0.168560 | Image2Text Acc: 91.67 | Text2Image Acc: 100.00 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:51 | INFO | Rank 0 | Global Steps: 430/5215 | Train Epoch: 1 [20640/250320 (8%)] | Loss: 0.365425 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:55 | INFO | Rank 0 | Global Steps: 440/5215 | Train Epoch: 1 [21120/250320 (8%)] | Loss: 0.561810 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:34:58 | INFO | Rank 0 | Global Steps: 450/5215 | Train Epoch: 1 [21600/250320 (9%)] | Loss: 0.176170 | Image2Text Acc: 91.67 | Text2Image Acc: 95.83 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:02 | INFO | Rank 0 | Global Steps: 460/5215 | Train Epoch: 1 [22080/250320 (9%)] | Loss: 0.339763 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:06 | INFO | Rank 0 | Global Steps: 470/5215 | Train Epoch: 1 [22560/250320 (9%)] | Loss: 0.495715 | Image2Text Acc: 89.58 | Text2Image Acc: 79.17 | Data Time: 0.023s | Batch Time: 0.365s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:09 | INFO | Rank 0 | Global Steps: 480/5215 | Train Epoch: 1 [23040/250320 (9%)] | Loss: 0.337739 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:13 | INFO | Rank 0 | Global Steps: 490/5215 | Train Epoch: 1 [23520/250320 (9%)] | Loss: 0.657292 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:17 | INFO | Rank 0 | Global Steps: 500/5215 | Train Epoch: 1 [24000/250320 (10%)] | Loss: 0.432810 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:20 | INFO | Rank 0 | Global Steps: 510/5215 | Train Epoch: 1 [24480/250320 (10%)] | Loss: 0.441927 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:24 | INFO | Rank 0 | Global Steps: 520/5215 | Train Epoch: 1 [24960/250320 (10%)] | Loss: 0.298030 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.018s | Batch Time: 0.375s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:28 | INFO | Rank 0 | Global Steps: 530/5215 | Train Epoch: 1 [25440/250320 (10%)] | Loss: 0.484386 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:31 | INFO | Rank 0 | Global Steps: 540/5215 | Train Epoch: 1 [25920/250320 (10%)] | Loss: 0.637108 | Image2Text Acc: 81.25 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:35 | INFO | Rank 0 | Global Steps: 550/5215 | Train Epoch: 1 [26400/250320 (11%)] | Loss: 0.214095 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:39 | INFO | Rank 0 | Global Steps: 560/5215 | Train Epoch: 1 [26880/250320 (11%)] | Loss: 0.358855 | Image2Text Acc: 87.50 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:42 | INFO | Rank 0 | Global Steps: 570/5215 | Train Epoch: 1 [27360/250320 (11%)] | Loss: 0.476662 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.021s | Batch Time: 0.368s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:46 | INFO | Rank 0 | Global Steps: 580/5215 | Train Epoch: 1 [27840/250320 (11%)] | Loss: 0.705014 | Image2Text Acc: 72.92 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.369s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:50 | INFO | Rank 0 | Global Steps: 590/5215 | Train Epoch: 1 [28320/250320 (11%)] | Loss: 0.389719 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.024s | Batch Time: 0.368s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:53 | INFO | Rank 0 | Global Steps: 600/5215 | Train Epoch: 1 [28800/250320 (12%)] | Loss: 0.567398 | Image2Text Acc: 81.25 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:35:57 | INFO | Rank 0 | Global Steps: 610/5215 | Train Epoch: 1 [29280/250320 (12%)] | Loss: 0.268593 | Image2Text Acc: 97.92 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.371s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:01 | INFO | Rank 0 | Global Steps: 620/5215 | Train Epoch: 1 [29760/250320 (12%)] | Loss: 0.450645 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.366s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:04 | INFO | Rank 0 | Global Steps: 630/5215 | Train Epoch: 1 [30240/250320 (12%)] | Loss: 0.451358 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:08 | INFO | Rank 0 | Global Steps: 640/5215 | Train Epoch: 1 [30720/250320 (12%)] | Loss: 0.474468 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.370s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:12 | INFO | Rank 0 | Global Steps: 650/5215 | Train Epoch: 1 [31200/250320 (12%)] | Loss: 0.540700 | Image2Text Acc: 83.33 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.394s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:16 | INFO | Rank 0 | Global Steps: 660/5215 | Train Epoch: 1 [31680/250320 (13%)] | Loss: 0.520677 | Image2Text Acc: 79.17 | Text2Image Acc: 79.17 | Data Time: 0.020s | Batch Time: 0.366s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:19 | INFO | Rank 0 | Global Steps: 670/5215 | Train Epoch: 1 [32160/250320 (13%)] | Loss: 0.156446 | Image2Text Acc: 97.92 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.366s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:23 | INFO | Rank 0 | Global Steps: 680/5215 | Train Epoch: 1 [32640/250320 (13%)] | Loss: 0.233609 | Image2Text Acc: 87.50 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:26 | INFO | Rank 0 | Global Steps: 690/5215 | Train Epoch: 1 [33120/250320 (13%)] | Loss: 0.453145 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.018s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:30 | INFO | Rank 0 | Global Steps: 700/5215 | Train Epoch: 1 [33600/250320 (13%)] | Loss: 0.429521 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:34 | INFO | Rank 0 | Global Steps: 710/5215 | Train Epoch: 1 [34080/250320 (14%)] | Loss: 0.699030 | Image2Text Acc: 81.25 | Text2Image Acc: 77.08 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:37 | INFO | Rank 0 | Global Steps: 720/5215 | Train Epoch: 1 [34560/250320 (14%)] | Loss: 0.258114 | Image2Text Acc: 95.83 | Text2Image Acc: 91.67 | Data Time: 0.022s | Batch Time: 0.373s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:41 | INFO | Rank 0 | Global Steps: 730/5215 | Train Epoch: 1 [35040/250320 (14%)] | Loss: 0.348272 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.018s | Batch Time: 0.378s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:45 | INFO | Rank 0 | Global Steps: 740/5215 | Train Epoch: 1 [35520/250320 (14%)] | Loss: 0.487886 | Image2Text Acc: 81.25 | Text2Image Acc: 81.25 | Data Time: 0.021s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:48 | INFO | Rank 0 | Global Steps: 750/5215 | Train Epoch: 1 [36000/250320 (14%)] | Loss: 0.333101 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.022s | Batch Time: 0.373s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:52 | INFO | Rank 0 | Global Steps: 760/5215 | Train Epoch: 1 [36480/250320 (15%)] | Loss: 0.126650 | Image2Text Acc: 95.83 | Text2Image Acc: 95.83 | Data Time: 0.018s | Batch Time: 0.360s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:36:56 | INFO | Rank 0 | Global Steps: 770/5215 | Train Epoch: 1 [36960/250320 (15%)] | Loss: 0.556339 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.018s | Batch Time: 0.365s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:00 | INFO | Rank 0 | Global Steps: 780/5215 | Train Epoch: 1 [37440/250320 (15%)] | Loss: 0.462434 | Image2Text Acc: 87.50 | Text2Image Acc: 81.25 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:03 | INFO | Rank 0 | Global Steps: 790/5215 | Train Epoch: 1 [37920/250320 (15%)] | Loss: 0.285763 | Image2Text Acc: 89.58 | Text2Image Acc: 95.83 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:07 | INFO | Rank 0 | Global Steps: 800/5215 | Train Epoch: 1 [38400/250320 (15%)] | Loss: 0.447099 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:11 | INFO | Rank 0 | Global Steps: 810/5215 | Train Epoch: 1 [38880/250320 (16%)] | Loss: 0.679671 | Image2Text Acc: 81.25 | Text2Image Acc: 79.17 | Data Time: 0.021s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:14 | INFO | Rank 0 | Global Steps: 820/5215 | Train Epoch: 1 [39360/250320 (16%)] | Loss: 0.190679 | Image2Text Acc: 97.92 | Text2Image Acc: 91.67 | Data Time: 0.023s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:18 | INFO | Rank 0 | Global Steps: 830/5215 | Train Epoch: 1 [39840/250320 (16%)] | Loss: 0.241572 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.023s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:22 | INFO | Rank 0 | Global Steps: 840/5215 | Train Epoch: 1 [40320/250320 (16%)] | Loss: 0.771389 | Image2Text Acc: 77.08 | Text2Image Acc: 77.08 | Data Time: 0.022s | Batch Time: 0.376s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:25 | INFO | Rank 0 | Global Steps: 850/5215 | Train Epoch: 1 [40800/250320 (16%)] | Loss: 0.292248 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:29 | INFO | Rank 0 | Global Steps: 860/5215 | Train Epoch: 1 [41280/250320 (16%)] | Loss: 0.325587 | Image2Text Acc: 85.42 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:33 | INFO | Rank 0 | Global Steps: 870/5215 | Train Epoch: 1 [41760/250320 (17%)] | Loss: 0.588850 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:36 | INFO | Rank 0 | Global Steps: 880/5215 | Train Epoch: 1 [42240/250320 (17%)] | Loss: 0.308825 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:40 | INFO | Rank 0 | Global Steps: 890/5215 | Train Epoch: 1 [42720/250320 (17%)] | Loss: 0.427337 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.038s | Batch Time: 0.393s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:44 | INFO | Rank 0 | Global Steps: 900/5215 | Train Epoch: 1 [43200/250320 (17%)] | Loss: 0.320133 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:47 | INFO | Rank 0 | Global Steps: 910/5215 | Train Epoch: 1 [43680/250320 (17%)] | Loss: 0.621858 | Image2Text Acc: 81.25 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:51 | INFO | Rank 0 | Global Steps: 920/5215 | Train Epoch: 1 [44160/250320 (18%)] | Loss: 0.216431 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:55 | INFO | Rank 0 | Global Steps: 930/5215 | Train Epoch: 1 [44640/250320 (18%)] | Loss: 0.303966 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:37:58 | INFO | Rank 0 | Global Steps: 940/5215 | Train Epoch: 1 [45120/250320 (18%)] | Loss: 0.490032 | Image2Text Acc: 91.67 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:38:02 | INFO | Rank 0 | Global Steps: 950/5215 | Train Epoch: 1 [45600/250320 (18%)] | Loss: 0.768142 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:38:06 | INFO | Rank 0 | Global Steps: 960/5215 | Train Epoch: 1 [46080/250320 (18%)] | Loss: 0.185704 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:38:09 | INFO | Rank 0 | Global Steps: 970/5215 | Train Epoch: 1 [46560/250320 (19%)] | Loss: 0.631737 | Image2Text Acc: 81.25 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:38:13 | INFO | Rank 0 | Global Steps: 980/5215 | Train Epoch: 1 [47040/250320 (19%)] | Loss: 0.734770 | Image2Text Acc: 87.50 | Text2Image Acc: 77.08 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:38:17 | INFO | Rank 0 | Global Steps: 990/5215 | Train Epoch: 1 [47520/250320 (19%)] | Loss: 0.383750 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.370s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:38:20 | INFO | Rank 0 | Global Steps: 1000/5215 | Train Epoch: 1 [48000/250320 (19%)] | Loss: 0.277359 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.022s | Batch Time: 0.376s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:38:20 | INFO | Rank 0 | Begin to eval on validation set (epoch 1 @ 1000 steps)...\n", + "/home/pai/lib/python3.6/site-packages/torch/utils/checkpoint.py:25: UserWarning: None of the inputs have requires_grad=True. Gradients will be None\n", + " warnings.warn(\"None of the inputs have requires_grad=True. Gradients will be None\")\n", + "2023-01-14,20:38:43 | INFO | Rank 0 | Evaluated 100/638 batches...\n", + "2023-01-14,20:39:06 | INFO | Rank 0 | Evaluated 200/638 batches...\n", + "2023-01-14,20:39:28 | INFO | Rank 0 | Evaluated 300/638 batches...\n", + "2023-01-14,20:39:50 | INFO | Rank 0 | Evaluated 400/638 batches...\n", + "2023-01-14,20:40:13 | INFO | Rank 0 | Evaluated 500/638 batches...\n", + "2023-01-14,20:40:35 | INFO | Rank 0 | Evaluated 600/638 batches...\n", + "2023-01-14,20:40:44 | INFO | Rank 0 | Validation Result (epoch 1 @ 1000 steps) | Valid Loss: 0.262804 | Image2Text Acc: 92.12 | Text2Image Acc: 91.45 | logit_scale: 4.605 | Valid Batch Size: 48\n", + "2023-01-14,20:40:48 | INFO | Rank 0 | Global Steps: 1010/5215 | Train Epoch: 1 [48480/250320 (19%)] | Loss: 0.353489 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:40:51 | INFO | Rank 0 | Global Steps: 1020/5215 | Train Epoch: 1 [48960/250320 (20%)] | Loss: 0.470332 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:40:55 | INFO | Rank 0 | Global Steps: 1030/5215 | Train Epoch: 1 [49440/250320 (20%)] | Loss: 0.329098 | Image2Text Acc: 89.58 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:40:59 | INFO | Rank 0 | Global Steps: 1040/5215 | Train Epoch: 1 [49920/250320 (20%)] | Loss: 0.525298 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.039s | Batch Time: 0.396s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:02 | INFO | Rank 0 | Global Steps: 1050/5215 | Train Epoch: 1 [50400/250320 (20%)] | Loss: 0.293415 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.025s | Batch Time: 0.370s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:06 | INFO | Rank 0 | Global Steps: 1060/5215 | Train Epoch: 1 [50880/250320 (20%)] | Loss: 0.474844 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:10 | INFO | Rank 0 | Global Steps: 1070/5215 | Train Epoch: 1 [51360/250320 (21%)] | Loss: 0.498422 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:13 | INFO | Rank 0 | Global Steps: 1080/5215 | Train Epoch: 1 [51840/250320 (21%)] | Loss: 0.426774 | Image2Text Acc: 83.33 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:17 | INFO | Rank 0 | Global Steps: 1090/5215 | Train Epoch: 1 [52320/250320 (21%)] | Loss: 0.513126 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.021s | Batch Time: 0.371s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:21 | INFO | Rank 0 | Global Steps: 1100/5215 | Train Epoch: 1 [52800/250320 (21%)] | Loss: 0.645229 | Image2Text Acc: 89.58 | Text2Image Acc: 81.25 | Data Time: 0.022s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:24 | INFO | Rank 0 | Global Steps: 1110/5215 | Train Epoch: 1 [53280/250320 (21%)] | Loss: 0.357841 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:28 | INFO | Rank 0 | Global Steps: 1120/5215 | Train Epoch: 1 [53760/250320 (21%)] | Loss: 0.467430 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:32 | INFO | Rank 0 | Global Steps: 1130/5215 | Train Epoch: 1 [54240/250320 (22%)] | Loss: 0.267649 | Image2Text Acc: 97.92 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.370s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:35 | INFO | Rank 0 | Global Steps: 1140/5215 | Train Epoch: 1 [54720/250320 (22%)] | Loss: 0.577349 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:39 | INFO | Rank 0 | Global Steps: 1150/5215 | Train Epoch: 1 [55200/250320 (22%)] | Loss: 0.521888 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.022s | Batch Time: 0.364s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:43 | INFO | Rank 0 | Global Steps: 1160/5215 | Train Epoch: 1 [55680/250320 (22%)] | Loss: 0.628832 | Image2Text Acc: 77.08 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:46 | INFO | Rank 0 | Global Steps: 1170/5215 | Train Epoch: 1 [56160/250320 (22%)] | Loss: 0.283388 | Image2Text Acc: 91.67 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:50 | INFO | Rank 0 | Global Steps: 1180/5215 | Train Epoch: 1 [56640/250320 (23%)] | Loss: 0.441064 | Image2Text Acc: 81.25 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:54 | INFO | Rank 0 | Global Steps: 1190/5215 | Train Epoch: 1 [57120/250320 (23%)] | Loss: 0.482407 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:41:57 | INFO | Rank 0 | Global Steps: 1200/5215 | Train Epoch: 1 [57600/250320 (23%)] | Loss: 0.190392 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.371s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:01 | INFO | Rank 0 | Global Steps: 1210/5215 | Train Epoch: 1 [58080/250320 (23%)] | Loss: 0.579661 | Image2Text Acc: 87.50 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:05 | INFO | Rank 0 | Global Steps: 1220/5215 | Train Epoch: 1 [58560/250320 (23%)] | Loss: 0.415587 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:08 | INFO | Rank 0 | Global Steps: 1230/5215 | Train Epoch: 1 [59040/250320 (24%)] | Loss: 0.524482 | Image2Text Acc: 83.33 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:12 | INFO | Rank 0 | Global Steps: 1240/5215 | Train Epoch: 1 [59520/250320 (24%)] | Loss: 0.429102 | Image2Text Acc: 89.58 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:16 | INFO | Rank 0 | Global Steps: 1250/5215 | Train Epoch: 1 [60000/250320 (24%)] | Loss: 0.335256 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:19 | INFO | Rank 0 | Global Steps: 1260/5215 | Train Epoch: 1 [60480/250320 (24%)] | Loss: 0.168375 | Image2Text Acc: 97.92 | Text2Image Acc: 97.92 | Data Time: 0.018s | Batch Time: 0.374s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:23 | INFO | Rank 0 | Global Steps: 1270/5215 | Train Epoch: 1 [60960/250320 (24%)] | Loss: 0.427466 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:27 | INFO | Rank 0 | Global Steps: 1280/5215 | Train Epoch: 1 [61440/250320 (25%)] | Loss: 0.421137 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.022s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:30 | INFO | Rank 0 | Global Steps: 1290/5215 | Train Epoch: 1 [61920/250320 (25%)] | Loss: 0.284803 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:34 | INFO | Rank 0 | Global Steps: 1300/5215 | Train Epoch: 1 [62400/250320 (25%)] | Loss: 0.544901 | Image2Text Acc: 79.17 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.373s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:38 | INFO | Rank 0 | Global Steps: 1310/5215 | Train Epoch: 1 [62880/250320 (25%)] | Loss: 0.278784 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:41 | INFO | Rank 0 | Global Steps: 1320/5215 | Train Epoch: 1 [63360/250320 (25%)] | Loss: 0.307508 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:45 | INFO | Rank 0 | Global Steps: 1330/5215 | Train Epoch: 1 [63840/250320 (26%)] | Loss: 0.500920 | Image2Text Acc: 81.25 | Text2Image Acc: 87.50 | Data Time: 0.018s | Batch Time: 0.366s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:49 | INFO | Rank 0 | Global Steps: 1340/5215 | Train Epoch: 1 [64320/250320 (26%)] | Loss: 0.482300 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.018s | Batch Time: 0.364s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:52 | INFO | Rank 0 | Global Steps: 1350/5215 | Train Epoch: 1 [64800/250320 (26%)] | Loss: 0.310206 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:42:56 | INFO | Rank 0 | Global Steps: 1360/5215 | Train Epoch: 1 [65280/250320 (26%)] | Loss: 0.353309 | Image2Text Acc: 95.83 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.369s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:00 | INFO | Rank 0 | Global Steps: 1370/5215 | Train Epoch: 1 [65760/250320 (26%)] | Loss: 0.635376 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.022s | Batch Time: 0.381s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:03 | INFO | Rank 0 | Global Steps: 1380/5215 | Train Epoch: 1 [66240/250320 (26%)] | Loss: 0.552023 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:07 | INFO | Rank 0 | Global Steps: 1390/5215 | Train Epoch: 1 [66720/250320 (27%)] | Loss: 0.468239 | Image2Text Acc: 83.33 | Text2Image Acc: 89.58 | Data Time: 0.022s | Batch Time: 0.372s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:11 | INFO | Rank 0 | Global Steps: 1400/5215 | Train Epoch: 1 [67200/250320 (27%)] | Loss: 0.669208 | Image2Text Acc: 83.33 | Text2Image Acc: 77.08 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:14 | INFO | Rank 0 | Global Steps: 1410/5215 | Train Epoch: 1 [67680/250320 (27%)] | Loss: 0.338586 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:18 | INFO | Rank 0 | Global Steps: 1420/5215 | Train Epoch: 1 [68160/250320 (27%)] | Loss: 0.526634 | Image2Text Acc: 79.17 | Text2Image Acc: 81.25 | Data Time: 0.021s | Batch Time: 0.364s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:22 | INFO | Rank 0 | Global Steps: 1430/5215 | Train Epoch: 1 [68640/250320 (27%)] | Loss: 0.365810 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:25 | INFO | Rank 0 | Global Steps: 1440/5215 | Train Epoch: 1 [69120/250320 (28%)] | Loss: 0.573479 | Image2Text Acc: 79.17 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.368s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:29 | INFO | Rank 0 | Global Steps: 1450/5215 | Train Epoch: 1 [69600/250320 (28%)] | Loss: 0.635550 | Image2Text Acc: 89.58 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.377s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:33 | INFO | Rank 0 | Global Steps: 1460/5215 | Train Epoch: 1 [70080/250320 (28%)] | Loss: 0.609029 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.021s | Batch Time: 0.363s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:36 | INFO | Rank 0 | Global Steps: 1470/5215 | Train Epoch: 1 [70560/250320 (28%)] | Loss: 0.549076 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.017s | Batch Time: 0.360s | LR: 0.000003 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:40 | INFO | Rank 0 | Global Steps: 1480/5215 | Train Epoch: 1 [71040/250320 (28%)] | Loss: 0.413826 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:44 | INFO | Rank 0 | Global Steps: 1490/5215 | Train Epoch: 1 [71520/250320 (29%)] | Loss: 0.297265 | Image2Text Acc: 93.75 | Text2Image Acc: 85.42 | Data Time: 0.022s | Batch Time: 0.372s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:47 | INFO | Rank 0 | Global Steps: 1500/5215 | Train Epoch: 1 [72000/250320 (29%)] | Loss: 0.553215 | Image2Text Acc: 87.50 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:51 | INFO | Rank 0 | Global Steps: 1510/5215 | Train Epoch: 1 [72480/250320 (29%)] | Loss: 0.610749 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:55 | INFO | Rank 0 | Global Steps: 1520/5215 | Train Epoch: 1 [72960/250320 (29%)] | Loss: 0.410754 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:43:58 | INFO | Rank 0 | Global Steps: 1530/5215 | Train Epoch: 1 [73440/250320 (29%)] | Loss: 0.491630 | Image2Text Acc: 81.25 | Text2Image Acc: 87.50 | Data Time: 0.026s | Batch Time: 0.368s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:02 | INFO | Rank 0 | Global Steps: 1540/5215 | Train Epoch: 1 [73920/250320 (30%)] | Loss: 0.182791 | Image2Text Acc: 97.92 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:06 | INFO | Rank 0 | Global Steps: 1550/5215 | Train Epoch: 1 [74400/250320 (30%)] | Loss: 0.398770 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.023s | Batch Time: 0.373s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:09 | INFO | Rank 0 | Global Steps: 1560/5215 | Train Epoch: 1 [74880/250320 (30%)] | Loss: 0.381658 | Image2Text Acc: 91.67 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:13 | INFO | Rank 0 | Global Steps: 1570/5215 | Train Epoch: 1 [75360/250320 (30%)] | Loss: 0.614655 | Image2Text Acc: 85.42 | Text2Image Acc: 81.25 | Data Time: 0.023s | Batch Time: 0.372s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:17 | INFO | Rank 0 | Global Steps: 1580/5215 | Train Epoch: 1 [75840/250320 (30%)] | Loss: 0.132453 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:20 | INFO | Rank 0 | Global Steps: 1590/5215 | Train Epoch: 1 [76320/250320 (30%)] | Loss: 0.488341 | Image2Text Acc: 81.25 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:24 | INFO | Rank 0 | Global Steps: 1600/5215 | Train Epoch: 1 [76800/250320 (31%)] | Loss: 0.377009 | Image2Text Acc: 85.42 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:28 | INFO | Rank 0 | Global Steps: 1610/5215 | Train Epoch: 1 [77280/250320 (31%)] | Loss: 0.183717 | Image2Text Acc: 95.83 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:31 | INFO | Rank 0 | Global Steps: 1620/5215 | Train Epoch: 1 [77760/250320 (31%)] | Loss: 0.456591 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.370s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:35 | INFO | Rank 0 | Global Steps: 1630/5215 | Train Epoch: 1 [78240/250320 (31%)] | Loss: 0.476531 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:39 | INFO | Rank 0 | Global Steps: 1640/5215 | Train Epoch: 1 [78720/250320 (31%)] | Loss: 0.452081 | Image2Text Acc: 79.17 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:42 | INFO | Rank 0 | Global Steps: 1650/5215 | Train Epoch: 1 [79200/250320 (32%)] | Loss: 0.321646 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.369s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:46 | INFO | Rank 0 | Global Steps: 1660/5215 | Train Epoch: 1 [79680/250320 (32%)] | Loss: 0.505794 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:50 | INFO | Rank 0 | Global Steps: 1670/5215 | Train Epoch: 1 [80160/250320 (32%)] | Loss: 0.267335 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:53 | INFO | Rank 0 | Global Steps: 1680/5215 | Train Epoch: 1 [80640/250320 (32%)] | Loss: 0.444049 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:44:57 | INFO | Rank 0 | Global Steps: 1690/5215 | Train Epoch: 1 [81120/250320 (32%)] | Loss: 0.385360 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:01 | INFO | Rank 0 | Global Steps: 1700/5215 | Train Epoch: 1 [81600/250320 (33%)] | Loss: 0.587886 | Image2Text Acc: 85.42 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:04 | INFO | Rank 0 | Global Steps: 1710/5215 | Train Epoch: 1 [82080/250320 (33%)] | Loss: 0.512576 | Image2Text Acc: 87.50 | Text2Image Acc: 81.25 | Data Time: 0.022s | Batch Time: 0.373s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:08 | INFO | Rank 0 | Global Steps: 1720/5215 | Train Epoch: 1 [82560/250320 (33%)] | Loss: 0.295488 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.367s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:12 | INFO | Rank 0 | Global Steps: 1730/5215 | Train Epoch: 1 [83040/250320 (33%)] | Loss: 0.208858 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.366s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:15 | INFO | Rank 0 | Global Steps: 1740/5215 | Train Epoch: 1 [83520/250320 (33%)] | Loss: 0.335569 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:19 | INFO | Rank 0 | Global Steps: 1750/5215 | Train Epoch: 1 [84000/250320 (34%)] | Loss: 0.487836 | Image2Text Acc: 83.33 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:23 | INFO | Rank 0 | Global Steps: 1760/5215 | Train Epoch: 1 [84480/250320 (34%)] | Loss: 0.120877 | Image2Text Acc: 93.75 | Text2Image Acc: 95.83 | Data Time: 0.023s | Batch Time: 0.365s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:26 | INFO | Rank 0 | Global Steps: 1770/5215 | Train Epoch: 1 [84960/250320 (34%)] | Loss: 0.314583 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:30 | INFO | Rank 0 | Global Steps: 1780/5215 | Train Epoch: 1 [85440/250320 (34%)] | Loss: 0.542514 | Image2Text Acc: 87.50 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:33 | INFO | Rank 0 | Global Steps: 1790/5215 | Train Epoch: 1 [85920/250320 (34%)] | Loss: 0.249195 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:37 | INFO | Rank 0 | Global Steps: 1800/5215 | Train Epoch: 1 [86400/250320 (35%)] | Loss: 0.240446 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.376s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:41 | INFO | Rank 0 | Global Steps: 1810/5215 | Train Epoch: 1 [86880/250320 (35%)] | Loss: 0.368746 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:44 | INFO | Rank 0 | Global Steps: 1820/5215 | Train Epoch: 1 [87360/250320 (35%)] | Loss: 0.278901 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.022s | Batch Time: 0.378s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:48 | INFO | Rank 0 | Global Steps: 1830/5215 | Train Epoch: 1 [87840/250320 (35%)] | Loss: 0.276916 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:52 | INFO | Rank 0 | Global Steps: 1840/5215 | Train Epoch: 1 [88320/250320 (35%)] | Loss: 0.216239 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:55 | INFO | Rank 0 | Global Steps: 1850/5215 | Train Epoch: 1 [88800/250320 (35%)] | Loss: 0.540644 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:45:59 | INFO | Rank 0 | Global Steps: 1860/5215 | Train Epoch: 1 [89280/250320 (36%)] | Loss: 0.429113 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:03 | INFO | Rank 0 | Global Steps: 1870/5215 | Train Epoch: 1 [89760/250320 (36%)] | Loss: 0.251663 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:06 | INFO | Rank 0 | Global Steps: 1880/5215 | Train Epoch: 1 [90240/250320 (36%)] | Loss: 0.526177 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:10 | INFO | Rank 0 | Global Steps: 1890/5215 | Train Epoch: 1 [90720/250320 (36%)] | Loss: 0.227536 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:14 | INFO | Rank 0 | Global Steps: 1900/5215 | Train Epoch: 1 [91200/250320 (36%)] | Loss: 0.671569 | Image2Text Acc: 77.08 | Text2Image Acc: 81.25 | Data Time: 0.018s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:17 | INFO | Rank 0 | Global Steps: 1910/5215 | Train Epoch: 1 [91680/250320 (37%)] | Loss: 0.428948 | Image2Text Acc: 81.25 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:21 | INFO | Rank 0 | Global Steps: 1920/5215 | Train Epoch: 1 [92160/250320 (37%)] | Loss: 0.332258 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:24 | INFO | Rank 0 | Global Steps: 1930/5215 | Train Epoch: 1 [92640/250320 (37%)] | Loss: 0.359012 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.376s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:28 | INFO | Rank 0 | Global Steps: 1940/5215 | Train Epoch: 1 [93120/250320 (37%)] | Loss: 0.802764 | Image2Text Acc: 77.08 | Text2Image Acc: 75.00 | Data Time: 0.022s | Batch Time: 0.373s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:32 | INFO | Rank 0 | Global Steps: 1950/5215 | Train Epoch: 1 [93600/250320 (37%)] | Loss: 0.622857 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:35 | INFO | Rank 0 | Global Steps: 1960/5215 | Train Epoch: 1 [94080/250320 (38%)] | Loss: 0.215016 | Image2Text Acc: 95.83 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:39 | INFO | Rank 0 | Global Steps: 1970/5215 | Train Epoch: 1 [94560/250320 (38%)] | Loss: 0.365139 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:43 | INFO | Rank 0 | Global Steps: 1980/5215 | Train Epoch: 1 [95040/250320 (38%)] | Loss: 0.214285 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.021s | Batch Time: 0.370s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:47 | INFO | Rank 0 | Global Steps: 1990/5215 | Train Epoch: 1 [95520/250320 (38%)] | Loss: 0.288544 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:50 | INFO | Rank 0 | Global Steps: 2000/5215 | Train Epoch: 1 [96000/250320 (38%)] | Loss: 0.508433 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:46:50 | INFO | Rank 0 | Begin to eval on validation set (epoch 1 @ 2000 steps)...\n", + "2023-01-14,20:47:13 | INFO | Rank 0 | Evaluated 100/638 batches...\n", + "2023-01-14,20:47:36 | INFO | Rank 0 | Evaluated 200/638 batches...\n", + "2023-01-14,20:47:58 | INFO | Rank 0 | Evaluated 300/638 batches...\n", + "2023-01-14,20:48:21 | INFO | Rank 0 | Evaluated 400/638 batches...\n", + "2023-01-14,20:48:43 | INFO | Rank 0 | Evaluated 500/638 batches...\n", + "2023-01-14,20:49:05 | INFO | Rank 0 | Evaluated 600/638 batches...\n", + "2023-01-14,20:49:14 | INFO | Rank 0 | Validation Result (epoch 1 @ 2000 steps) | Valid Loss: 0.253930 | Image2Text Acc: 92.30 | Text2Image Acc: 91.60 | logit_scale: 4.605 | Valid Batch Size: 48\n", + "2023-01-14,20:49:18 | INFO | Rank 0 | Global Steps: 2010/5215 | Train Epoch: 1 [96480/250320 (39%)] | Loss: 0.393346 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:22 | INFO | Rank 0 | Global Steps: 2020/5215 | Train Epoch: 1 [96960/250320 (39%)] | Loss: 0.236084 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.022s | Batch Time: 0.373s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:25 | INFO | Rank 0 | Global Steps: 2030/5215 | Train Epoch: 1 [97440/250320 (39%)] | Loss: 0.247176 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:29 | INFO | Rank 0 | Global Steps: 2040/5215 | Train Epoch: 1 [97920/250320 (39%)] | Loss: 0.489473 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.023s | Batch Time: 0.366s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:33 | INFO | Rank 0 | Global Steps: 2050/5215 | Train Epoch: 1 [98400/250320 (39%)] | Loss: 0.177373 | Image2Text Acc: 93.75 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:37 | INFO | Rank 0 | Global Steps: 2060/5215 | Train Epoch: 1 [98880/250320 (40%)] | Loss: 0.446684 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:40 | INFO | Rank 0 | Global Steps: 2070/5215 | Train Epoch: 1 [99360/250320 (40%)] | Loss: 0.361537 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:44 | INFO | Rank 0 | Global Steps: 2080/5215 | Train Epoch: 1 [99840/250320 (40%)] | Loss: 0.532427 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:47 | INFO | Rank 0 | Global Steps: 2090/5215 | Train Epoch: 1 [100320/250320 (40%)] | Loss: 0.338953 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:51 | INFO | Rank 0 | Global Steps: 2100/5215 | Train Epoch: 1 [100800/250320 (40%)] | Loss: 0.324819 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:55 | INFO | Rank 0 | Global Steps: 2110/5215 | Train Epoch: 1 [101280/250320 (40%)] | Loss: 0.305668 | Image2Text Acc: 95.83 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:49:58 | INFO | Rank 0 | Global Steps: 2120/5215 | Train Epoch: 1 [101760/250320 (41%)] | Loss: 0.428335 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:02 | INFO | Rank 0 | Global Steps: 2130/5215 | Train Epoch: 1 [102240/250320 (41%)] | Loss: 0.235824 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.025s | Batch Time: 0.381s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:06 | INFO | Rank 0 | Global Steps: 2140/5215 | Train Epoch: 1 [102720/250320 (41%)] | Loss: 0.396628 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:09 | INFO | Rank 0 | Global Steps: 2150/5215 | Train Epoch: 1 [103200/250320 (41%)] | Loss: 0.435109 | Image2Text Acc: 83.33 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:13 | INFO | Rank 0 | Global Steps: 2160/5215 | Train Epoch: 1 [103680/250320 (41%)] | Loss: 0.549152 | Image2Text Acc: 87.50 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:17 | INFO | Rank 0 | Global Steps: 2170/5215 | Train Epoch: 1 [104160/250320 (42%)] | Loss: 0.462073 | Image2Text Acc: 91.67 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.376s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:20 | INFO | Rank 0 | Global Steps: 2180/5215 | Train Epoch: 1 [104640/250320 (42%)] | Loss: 0.337940 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.022s | Batch Time: 0.365s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:24 | INFO | Rank 0 | Global Steps: 2190/5215 | Train Epoch: 1 [105120/250320 (42%)] | Loss: 0.358022 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.382s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:28 | INFO | Rank 0 | Global Steps: 2200/5215 | Train Epoch: 1 [105600/250320 (42%)] | Loss: 0.526386 | Image2Text Acc: 79.17 | Text2Image Acc: 83.33 | Data Time: 0.022s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:32 | INFO | Rank 0 | Global Steps: 2210/5215 | Train Epoch: 1 [106080/250320 (42%)] | Loss: 0.232042 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.380s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:35 | INFO | Rank 0 | Global Steps: 2220/5215 | Train Epoch: 1 [106560/250320 (43%)] | Loss: 0.510376 | Image2Text Acc: 85.42 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:39 | INFO | Rank 0 | Global Steps: 2230/5215 | Train Epoch: 1 [107040/250320 (43%)] | Loss: 0.750720 | Image2Text Acc: 79.17 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:43 | INFO | Rank 0 | Global Steps: 2240/5215 | Train Epoch: 1 [107520/250320 (43%)] | Loss: 0.297693 | Image2Text Acc: 93.75 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:46 | INFO | Rank 0 | Global Steps: 2250/5215 | Train Epoch: 1 [108000/250320 (43%)] | Loss: 0.241137 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.366s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:50 | INFO | Rank 0 | Global Steps: 2260/5215 | Train Epoch: 1 [108480/250320 (43%)] | Loss: 0.691053 | Image2Text Acc: 79.17 | Text2Image Acc: 77.08 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:53 | INFO | Rank 0 | Global Steps: 2270/5215 | Train Epoch: 1 [108960/250320 (44%)] | Loss: 0.412174 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.021s | Batch Time: 0.367s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:50:57 | INFO | Rank 0 | Global Steps: 2280/5215 | Train Epoch: 1 [109440/250320 (44%)] | Loss: 0.310207 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:01 | INFO | Rank 0 | Global Steps: 2290/5215 | Train Epoch: 1 [109920/250320 (44%)] | Loss: 0.342477 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:04 | INFO | Rank 0 | Global Steps: 2300/5215 | Train Epoch: 1 [110400/250320 (44%)] | Loss: 0.565625 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:08 | INFO | Rank 0 | Global Steps: 2310/5215 | Train Epoch: 1 [110880/250320 (44%)] | Loss: 0.341602 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:12 | INFO | Rank 0 | Global Steps: 2320/5215 | Train Epoch: 1 [111360/250320 (44%)] | Loss: 0.529862 | Image2Text Acc: 85.42 | Text2Image Acc: 75.00 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:15 | INFO | Rank 0 | Global Steps: 2330/5215 | Train Epoch: 1 [111840/250320 (45%)] | Loss: 0.318515 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.374s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:19 | INFO | Rank 0 | Global Steps: 2340/5215 | Train Epoch: 1 [112320/250320 (45%)] | Loss: 0.222913 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.025s | Batch Time: 0.367s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:23 | INFO | Rank 0 | Global Steps: 2350/5215 | Train Epoch: 1 [112800/250320 (45%)] | Loss: 0.673159 | Image2Text Acc: 79.17 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:27 | INFO | Rank 0 | Global Steps: 2360/5215 | Train Epoch: 1 [113280/250320 (45%)] | Loss: 0.820517 | Image2Text Acc: 77.08 | Text2Image Acc: 81.25 | Data Time: 0.023s | Batch Time: 0.377s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:30 | INFO | Rank 0 | Global Steps: 2370/5215 | Train Epoch: 1 [113760/250320 (45%)] | Loss: 0.331657 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.376s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:34 | INFO | Rank 0 | Global Steps: 2380/5215 | Train Epoch: 1 [114240/250320 (46%)] | Loss: 0.300309 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:37 | INFO | Rank 0 | Global Steps: 2390/5215 | Train Epoch: 1 [114720/250320 (46%)] | Loss: 0.251794 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:41 | INFO | Rank 0 | Global Steps: 2400/5215 | Train Epoch: 1 [115200/250320 (46%)] | Loss: 0.228814 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:45 | INFO | Rank 0 | Global Steps: 2410/5215 | Train Epoch: 1 [115680/250320 (46%)] | Loss: 0.579475 | Image2Text Acc: 85.42 | Text2Image Acc: 77.08 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:48 | INFO | Rank 0 | Global Steps: 2420/5215 | Train Epoch: 1 [116160/250320 (46%)] | Loss: 0.268366 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.018s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:52 | INFO | Rank 0 | Global Steps: 2430/5215 | Train Epoch: 1 [116640/250320 (47%)] | Loss: 0.392079 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.021s | Batch Time: 0.371s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:56 | INFO | Rank 0 | Global Steps: 2440/5215 | Train Epoch: 1 [117120/250320 (47%)] | Loss: 0.324616 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:51:59 | INFO | Rank 0 | Global Steps: 2450/5215 | Train Epoch: 1 [117600/250320 (47%)] | Loss: 0.481112 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:03 | INFO | Rank 0 | Global Steps: 2460/5215 | Train Epoch: 1 [118080/250320 (47%)] | Loss: 0.431430 | Image2Text Acc: 87.50 | Text2Image Acc: 83.33 | Data Time: 0.018s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:07 | INFO | Rank 0 | Global Steps: 2470/5215 | Train Epoch: 1 [118560/250320 (47%)] | Loss: 0.092991 | Image2Text Acc: 97.92 | Text2Image Acc: 95.83 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:10 | INFO | Rank 0 | Global Steps: 2480/5215 | Train Epoch: 1 [119040/250320 (48%)] | Loss: 0.246772 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.022s | Batch Time: 0.371s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:14 | INFO | Rank 0 | Global Steps: 2490/5215 | Train Epoch: 1 [119520/250320 (48%)] | Loss: 0.489620 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:18 | INFO | Rank 0 | Global Steps: 2500/5215 | Train Epoch: 1 [120000/250320 (48%)] | Loss: 0.644723 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:21 | INFO | Rank 0 | Global Steps: 2510/5215 | Train Epoch: 1 [120480/250320 (48%)] | Loss: 0.325605 | Image2Text Acc: 93.75 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:25 | INFO | Rank 0 | Global Steps: 2520/5215 | Train Epoch: 1 [120960/250320 (48%)] | Loss: 0.304364 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.023s | Batch Time: 0.365s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:29 | INFO | Rank 0 | Global Steps: 2530/5215 | Train Epoch: 1 [121440/250320 (49%)] | Loss: 0.366834 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.386s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:33 | INFO | Rank 0 | Global Steps: 2540/5215 | Train Epoch: 1 [121920/250320 (49%)] | Loss: 0.383688 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.023s | Batch Time: 0.379s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:36 | INFO | Rank 0 | Global Steps: 2550/5215 | Train Epoch: 1 [122400/250320 (49%)] | Loss: 0.205350 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:40 | INFO | Rank 0 | Global Steps: 2560/5215 | Train Epoch: 1 [122880/250320 (49%)] | Loss: 0.536348 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:43 | INFO | Rank 0 | Global Steps: 2570/5215 | Train Epoch: 1 [123360/250320 (49%)] | Loss: 0.297162 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:47 | INFO | Rank 0 | Global Steps: 2580/5215 | Train Epoch: 1 [123840/250320 (49%)] | Loss: 0.297163 | Image2Text Acc: 93.75 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.372s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:51 | INFO | Rank 0 | Global Steps: 2590/5215 | Train Epoch: 1 [124320/250320 (50%)] | Loss: 0.352722 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.027s | Batch Time: 0.372s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:55 | INFO | Rank 0 | Global Steps: 2600/5215 | Train Epoch: 1 [124800/250320 (50%)] | Loss: 0.496191 | Image2Text Acc: 83.33 | Text2Image Acc: 75.00 | Data Time: 0.020s | Batch Time: 0.364s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:52:58 | INFO | Rank 0 | Global Steps: 2610/5215 | Train Epoch: 1 [125280/250320 (50%)] | Loss: 0.472608 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:02 | INFO | Rank 0 | Global Steps: 2620/5215 | Train Epoch: 1 [125760/250320 (50%)] | Loss: 0.373154 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.366s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:05 | INFO | Rank 0 | Global Steps: 2630/5215 | Train Epoch: 1 [126240/250320 (50%)] | Loss: 0.407097 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:09 | INFO | Rank 0 | Global Steps: 2640/5215 | Train Epoch: 1 [126720/250320 (51%)] | Loss: 0.259960 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:13 | INFO | Rank 0 | Global Steps: 2650/5215 | Train Epoch: 1 [127200/250320 (51%)] | Loss: 0.287041 | Image2Text Acc: 95.83 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.373s | LR: 0.000002 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:16 | INFO | Rank 0 | Global Steps: 2660/5215 | Train Epoch: 1 [127680/250320 (51%)] | Loss: 0.281949 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:20 | INFO | Rank 0 | Global Steps: 2670/5215 | Train Epoch: 1 [128160/250320 (51%)] | Loss: 0.483760 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.021s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:24 | INFO | Rank 0 | Global Steps: 2680/5215 | Train Epoch: 1 [128640/250320 (51%)] | Loss: 0.249935 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:27 | INFO | Rank 0 | Global Steps: 2690/5215 | Train Epoch: 1 [129120/250320 (52%)] | Loss: 0.228626 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:31 | INFO | Rank 0 | Global Steps: 2700/5215 | Train Epoch: 1 [129600/250320 (52%)] | Loss: 0.411693 | Image2Text Acc: 85.42 | Text2Image Acc: 89.58 | Data Time: 0.021s | Batch Time: 0.372s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:35 | INFO | Rank 0 | Global Steps: 2710/5215 | Train Epoch: 1 [130080/250320 (52%)] | Loss: 0.493524 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.022s | Batch Time: 0.371s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:38 | INFO | Rank 0 | Global Steps: 2720/5215 | Train Epoch: 1 [130560/250320 (52%)] | Loss: 0.219654 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:42 | INFO | Rank 0 | Global Steps: 2730/5215 | Train Epoch: 1 [131040/250320 (52%)] | Loss: 0.468917 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:46 | INFO | Rank 0 | Global Steps: 2740/5215 | Train Epoch: 1 [131520/250320 (53%)] | Loss: 0.251967 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.018s | Batch Time: 0.360s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:49 | INFO | Rank 0 | Global Steps: 2750/5215 | Train Epoch: 1 [132000/250320 (53%)] | Loss: 0.437946 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.365s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:53 | INFO | Rank 0 | Global Steps: 2760/5215 | Train Epoch: 1 [132480/250320 (53%)] | Loss: 0.753715 | Image2Text Acc: 75.00 | Text2Image Acc: 77.08 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:53:57 | INFO | Rank 0 | Global Steps: 2770/5215 | Train Epoch: 1 [132960/250320 (53%)] | Loss: 0.520218 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.366s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:00 | INFO | Rank 0 | Global Steps: 2780/5215 | Train Epoch: 1 [133440/250320 (53%)] | Loss: 0.436514 | Image2Text Acc: 89.58 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.368s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:04 | INFO | Rank 0 | Global Steps: 2790/5215 | Train Epoch: 1 [133920/250320 (53%)] | Loss: 0.319248 | Image2Text Acc: 87.50 | Text2Image Acc: 93.75 | Data Time: 0.022s | Batch Time: 0.372s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:08 | INFO | Rank 0 | Global Steps: 2800/5215 | Train Epoch: 1 [134400/250320 (54%)] | Loss: 0.147926 | Image2Text Acc: 95.83 | Text2Image Acc: 95.83 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:11 | INFO | Rank 0 | Global Steps: 2810/5215 | Train Epoch: 1 [134880/250320 (54%)] | Loss: 0.329935 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:15 | INFO | Rank 0 | Global Steps: 2820/5215 | Train Epoch: 1 [135360/250320 (54%)] | Loss: 0.409039 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.018s | Batch Time: 0.375s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:19 | INFO | Rank 0 | Global Steps: 2830/5215 | Train Epoch: 1 [135840/250320 (54%)] | Loss: 0.349650 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.021s | Batch Time: 0.365s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:22 | INFO | Rank 0 | Global Steps: 2840/5215 | Train Epoch: 1 [136320/250320 (54%)] | Loss: 0.207253 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:26 | INFO | Rank 0 | Global Steps: 2850/5215 | Train Epoch: 1 [136800/250320 (55%)] | Loss: 0.643075 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:30 | INFO | Rank 0 | Global Steps: 2860/5215 | Train Epoch: 1 [137280/250320 (55%)] | Loss: 0.235392 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:33 | INFO | Rank 0 | Global Steps: 2870/5215 | Train Epoch: 1 [137760/250320 (55%)] | Loss: 0.228034 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:37 | INFO | Rank 0 | Global Steps: 2880/5215 | Train Epoch: 1 [138240/250320 (55%)] | Loss: 0.245669 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:41 | INFO | Rank 0 | Global Steps: 2890/5215 | Train Epoch: 1 [138720/250320 (55%)] | Loss: 0.166860 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:44 | INFO | Rank 0 | Global Steps: 2900/5215 | Train Epoch: 1 [139200/250320 (56%)] | Loss: 0.338099 | Image2Text Acc: 91.67 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:48 | INFO | Rank 0 | Global Steps: 2910/5215 | Train Epoch: 1 [139680/250320 (56%)] | Loss: 0.370300 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:52 | INFO | Rank 0 | Global Steps: 2920/5215 | Train Epoch: 1 [140160/250320 (56%)] | Loss: 0.400969 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:55 | INFO | Rank 0 | Global Steps: 2930/5215 | Train Epoch: 1 [140640/250320 (56%)] | Loss: 0.455307 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:54:59 | INFO | Rank 0 | Global Steps: 2940/5215 | Train Epoch: 1 [141120/250320 (56%)] | Loss: 0.533848 | Image2Text Acc: 87.50 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:55:03 | INFO | Rank 0 | Global Steps: 2950/5215 | Train Epoch: 1 [141600/250320 (57%)] | Loss: 0.435563 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.018s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:55:06 | INFO | Rank 0 | Global Steps: 2960/5215 | Train Epoch: 1 [142080/250320 (57%)] | Loss: 0.538444 | Image2Text Acc: 89.58 | Text2Image Acc: 83.33 | Data Time: 0.018s | Batch Time: 0.369s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:55:10 | INFO | Rank 0 | Global Steps: 2970/5215 | Train Epoch: 1 [142560/250320 (57%)] | Loss: 0.484170 | Image2Text Acc: 79.17 | Text2Image Acc: 83.33 | Data Time: 0.018s | Batch Time: 0.366s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:55:14 | INFO | Rank 0 | Global Steps: 2980/5215 | Train Epoch: 1 [143040/250320 (57%)] | Loss: 0.279289 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:55:17 | INFO | Rank 0 | Global Steps: 2990/5215 | Train Epoch: 1 [143520/250320 (57%)] | Loss: 0.269024 | Image2Text Acc: 85.42 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:55:21 | INFO | Rank 0 | Global Steps: 3000/5215 | Train Epoch: 1 [144000/250320 (58%)] | Loss: 0.299232 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.023s | Batch Time: 0.379s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:55:21 | INFO | Rank 0 | Begin to eval on validation set (epoch 1 @ 3000 steps)...\n", + "2023-01-14,20:55:44 | INFO | Rank 0 | Evaluated 100/638 batches...\n", + "2023-01-14,20:56:06 | INFO | Rank 0 | Evaluated 200/638 batches...\n", + "2023-01-14,20:56:29 | INFO | Rank 0 | Evaluated 300/638 batches...\n", + "2023-01-14,20:56:51 | INFO | Rank 0 | Evaluated 400/638 batches...\n", + "2023-01-14,20:57:14 | INFO | Rank 0 | Evaluated 500/638 batches...\n", + "2023-01-14,20:57:36 | INFO | Rank 0 | Evaluated 600/638 batches...\n", + "2023-01-14,20:57:45 | INFO | Rank 0 | Validation Result (epoch 1 @ 3000 steps) | Valid Loss: 0.245162 | Image2Text Acc: 92.51 | Text2Image Acc: 91.86 | logit_scale: 4.605 | Valid Batch Size: 48\n", + "2023-01-14,20:57:49 | INFO | Rank 0 | Global Steps: 3010/5215 | Train Epoch: 1 [144480/250320 (58%)] | Loss: 0.272546 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:57:52 | INFO | Rank 0 | Global Steps: 3020/5215 | Train Epoch: 1 [144960/250320 (58%)] | Loss: 0.447747 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.018s | Batch Time: 0.360s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:57:56 | INFO | Rank 0 | Global Steps: 3030/5215 | Train Epoch: 1 [145440/250320 (58%)] | Loss: 0.476081 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.021s | Batch Time: 0.376s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:00 | INFO | Rank 0 | Global Steps: 3040/5215 | Train Epoch: 1 [145920/250320 (58%)] | Loss: 0.267517 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:03 | INFO | Rank 0 | Global Steps: 3050/5215 | Train Epoch: 1 [146400/250320 (58%)] | Loss: 0.226274 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.372s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:07 | INFO | Rank 0 | Global Steps: 3060/5215 | Train Epoch: 1 [146880/250320 (59%)] | Loss: 0.217604 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:11 | INFO | Rank 0 | Global Steps: 3070/5215 | Train Epoch: 1 [147360/250320 (59%)] | Loss: 0.177885 | Image2Text Acc: 95.83 | Text2Image Acc: 91.67 | Data Time: 0.022s | Batch Time: 0.373s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:14 | INFO | Rank 0 | Global Steps: 3080/5215 | Train Epoch: 1 [147840/250320 (59%)] | Loss: 0.498376 | Image2Text Acc: 79.17 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:18 | INFO | Rank 0 | Global Steps: 3090/5215 | Train Epoch: 1 [148320/250320 (59%)] | Loss: 0.330423 | Image2Text Acc: 89.58 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:22 | INFO | Rank 0 | Global Steps: 3100/5215 | Train Epoch: 1 [148800/250320 (59%)] | Loss: 0.256121 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:25 | INFO | Rank 0 | Global Steps: 3110/5215 | Train Epoch: 1 [149280/250320 (60%)] | Loss: 0.385000 | Image2Text Acc: 93.75 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:29 | INFO | Rank 0 | Global Steps: 3120/5215 | Train Epoch: 1 [149760/250320 (60%)] | Loss: 0.430297 | Image2Text Acc: 93.75 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:33 | INFO | Rank 0 | Global Steps: 3130/5215 | Train Epoch: 1 [150240/250320 (60%)] | Loss: 0.522045 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.024s | Batch Time: 0.367s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:36 | INFO | Rank 0 | Global Steps: 3140/5215 | Train Epoch: 1 [150720/250320 (60%)] | Loss: 0.290708 | Image2Text Acc: 93.75 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:40 | INFO | Rank 0 | Global Steps: 3150/5215 | Train Epoch: 1 [151200/250320 (60%)] | Loss: 0.302918 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.022s | Batch Time: 0.373s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:44 | INFO | Rank 0 | Global Steps: 3160/5215 | Train Epoch: 1 [151680/250320 (61%)] | Loss: 0.187307 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.023s | Batch Time: 0.373s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:47 | INFO | Rank 0 | Global Steps: 3170/5215 | Train Epoch: 1 [152160/250320 (61%)] | Loss: 0.388520 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:51 | INFO | Rank 0 | Global Steps: 3180/5215 | Train Epoch: 1 [152640/250320 (61%)] | Loss: 0.156572 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.369s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:55 | INFO | Rank 0 | Global Steps: 3190/5215 | Train Epoch: 1 [153120/250320 (61%)] | Loss: 0.347776 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:58:58 | INFO | Rank 0 | Global Steps: 3200/5215 | Train Epoch: 1 [153600/250320 (61%)] | Loss: 0.381887 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:02 | INFO | Rank 0 | Global Steps: 3210/5215 | Train Epoch: 1 [154080/250320 (62%)] | Loss: 0.524301 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.371s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:06 | INFO | Rank 0 | Global Steps: 3220/5215 | Train Epoch: 1 [154560/250320 (62%)] | Loss: 0.297189 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.022s | Batch Time: 0.374s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:09 | INFO | Rank 0 | Global Steps: 3230/5215 | Train Epoch: 1 [155040/250320 (62%)] | Loss: 0.461027 | Image2Text Acc: 85.42 | Text2Image Acc: 81.25 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:13 | INFO | Rank 0 | Global Steps: 3240/5215 | Train Epoch: 1 [155520/250320 (62%)] | Loss: 0.436980 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:17 | INFO | Rank 0 | Global Steps: 3250/5215 | Train Epoch: 1 [156000/250320 (62%)] | Loss: 0.403678 | Image2Text Acc: 85.42 | Text2Image Acc: 81.25 | Data Time: 0.018s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:20 | INFO | Rank 0 | Global Steps: 3260/5215 | Train Epoch: 1 [156480/250320 (63%)] | Loss: 0.213571 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.021s | Batch Time: 0.375s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:24 | INFO | Rank 0 | Global Steps: 3270/5215 | Train Epoch: 1 [156960/250320 (63%)] | Loss: 0.356295 | Image2Text Acc: 87.50 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:28 | INFO | Rank 0 | Global Steps: 3280/5215 | Train Epoch: 1 [157440/250320 (63%)] | Loss: 0.166876 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:31 | INFO | Rank 0 | Global Steps: 3290/5215 | Train Epoch: 1 [157920/250320 (63%)] | Loss: 0.092873 | Image2Text Acc: 97.92 | Text2Image Acc: 97.92 | Data Time: 0.022s | Batch Time: 0.372s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:35 | INFO | Rank 0 | Global Steps: 3300/5215 | Train Epoch: 1 [158400/250320 (63%)] | Loss: 0.378582 | Image2Text Acc: 89.58 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:39 | INFO | Rank 0 | Global Steps: 3310/5215 | Train Epoch: 1 [158880/250320 (63%)] | Loss: 0.335712 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:42 | INFO | Rank 0 | Global Steps: 3320/5215 | Train Epoch: 1 [159360/250320 (64%)] | Loss: 0.232245 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:46 | INFO | Rank 0 | Global Steps: 3330/5215 | Train Epoch: 1 [159840/250320 (64%)] | Loss: 0.463921 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.367s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:50 | INFO | Rank 0 | Global Steps: 3340/5215 | Train Epoch: 1 [160320/250320 (64%)] | Loss: 0.489526 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:53 | INFO | Rank 0 | Global Steps: 3350/5215 | Train Epoch: 1 [160800/250320 (64%)] | Loss: 0.475601 | Image2Text Acc: 81.25 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,20:59:57 | INFO | Rank 0 | Global Steps: 3360/5215 | Train Epoch: 1 [161280/250320 (64%)] | Loss: 0.423765 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.369s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:01 | INFO | Rank 0 | Global Steps: 3370/5215 | Train Epoch: 1 [161760/250320 (65%)] | Loss: 0.205251 | Image2Text Acc: 87.50 | Text2Image Acc: 97.92 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:04 | INFO | Rank 0 | Global Steps: 3380/5215 | Train Epoch: 1 [162240/250320 (65%)] | Loss: 0.499233 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:08 | INFO | Rank 0 | Global Steps: 3390/5215 | Train Epoch: 1 [162720/250320 (65%)] | Loss: 0.453877 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:12 | INFO | Rank 0 | Global Steps: 3400/5215 | Train Epoch: 1 [163200/250320 (65%)] | Loss: 0.205265 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:15 | INFO | Rank 0 | Global Steps: 3410/5215 | Train Epoch: 1 [163680/250320 (65%)] | Loss: 0.492467 | Image2Text Acc: 83.33 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.367s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:19 | INFO | Rank 0 | Global Steps: 3420/5215 | Train Epoch: 1 [164160/250320 (66%)] | Loss: 0.634184 | Image2Text Acc: 85.42 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.370s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:23 | INFO | Rank 0 | Global Steps: 3430/5215 | Train Epoch: 1 [164640/250320 (66%)] | Loss: 0.384186 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:26 | INFO | Rank 0 | Global Steps: 3440/5215 | Train Epoch: 1 [165120/250320 (66%)] | Loss: 0.489579 | Image2Text Acc: 87.50 | Text2Image Acc: 77.08 | Data Time: 0.023s | Batch Time: 0.365s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:30 | INFO | Rank 0 | Global Steps: 3450/5215 | Train Epoch: 1 [165600/250320 (66%)] | Loss: 0.299198 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:34 | INFO | Rank 0 | Global Steps: 3460/5215 | Train Epoch: 1 [166080/250320 (66%)] | Loss: 0.154600 | Image2Text Acc: 95.83 | Text2Image Acc: 97.92 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:37 | INFO | Rank 0 | Global Steps: 3470/5215 | Train Epoch: 1 [166560/250320 (67%)] | Loss: 0.248392 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:41 | INFO | Rank 0 | Global Steps: 3480/5215 | Train Epoch: 1 [167040/250320 (67%)] | Loss: 0.321048 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:45 | INFO | Rank 0 | Global Steps: 3490/5215 | Train Epoch: 1 [167520/250320 (67%)] | Loss: 0.302907 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:48 | INFO | Rank 0 | Global Steps: 3500/5215 | Train Epoch: 1 [168000/250320 (67%)] | Loss: 0.296233 | Image2Text Acc: 95.83 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:52 | INFO | Rank 0 | Global Steps: 3510/5215 | Train Epoch: 1 [168480/250320 (67%)] | Loss: 0.310646 | Image2Text Acc: 91.67 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:56 | INFO | Rank 0 | Global Steps: 3520/5215 | Train Epoch: 1 [168960/250320 (67%)] | Loss: 0.258731 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:00:59 | INFO | Rank 0 | Global Steps: 3530/5215 | Train Epoch: 1 [169440/250320 (68%)] | Loss: 0.422833 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:03 | INFO | Rank 0 | Global Steps: 3540/5215 | Train Epoch: 1 [169920/250320 (68%)] | Loss: 0.416619 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:07 | INFO | Rank 0 | Global Steps: 3550/5215 | Train Epoch: 1 [170400/250320 (68%)] | Loss: 0.758026 | Image2Text Acc: 75.00 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:10 | INFO | Rank 0 | Global Steps: 3560/5215 | Train Epoch: 1 [170880/250320 (68%)] | Loss: 0.379954 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.018s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:14 | INFO | Rank 0 | Global Steps: 3570/5215 | Train Epoch: 1 [171360/250320 (68%)] | Loss: 0.478845 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:17 | INFO | Rank 0 | Global Steps: 3580/5215 | Train Epoch: 1 [171840/250320 (69%)] | Loss: 0.339050 | Image2Text Acc: 87.50 | Text2Image Acc: 81.25 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:21 | INFO | Rank 0 | Global Steps: 3590/5215 | Train Epoch: 1 [172320/250320 (69%)] | Loss: 0.687952 | Image2Text Acc: 77.08 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:25 | INFO | Rank 0 | Global Steps: 3600/5215 | Train Epoch: 1 [172800/250320 (69%)] | Loss: 0.493248 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:28 | INFO | Rank 0 | Global Steps: 3610/5215 | Train Epoch: 1 [173280/250320 (69%)] | Loss: 0.226441 | Image2Text Acc: 91.67 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:32 | INFO | Rank 0 | Global Steps: 3620/5215 | Train Epoch: 1 [173760/250320 (69%)] | Loss: 0.383224 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:36 | INFO | Rank 0 | Global Steps: 3630/5215 | Train Epoch: 1 [174240/250320 (70%)] | Loss: 0.515392 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:39 | INFO | Rank 0 | Global Steps: 3640/5215 | Train Epoch: 1 [174720/250320 (70%)] | Loss: 0.308256 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:43 | INFO | Rank 0 | Global Steps: 3650/5215 | Train Epoch: 1 [175200/250320 (70%)] | Loss: 0.703550 | Image2Text Acc: 79.17 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:46 | INFO | Rank 0 | Global Steps: 3660/5215 | Train Epoch: 1 [175680/250320 (70%)] | Loss: 0.371586 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:50 | INFO | Rank 0 | Global Steps: 3670/5215 | Train Epoch: 1 [176160/250320 (70%)] | Loss: 0.078936 | Image2Text Acc: 97.92 | Text2Image Acc: 97.92 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:54 | INFO | Rank 0 | Global Steps: 3680/5215 | Train Epoch: 1 [176640/250320 (71%)] | Loss: 0.286770 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:01:57 | INFO | Rank 0 | Global Steps: 3690/5215 | Train Epoch: 1 [177120/250320 (71%)] | Loss: 0.241136 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:01 | INFO | Rank 0 | Global Steps: 3700/5215 | Train Epoch: 1 [177600/250320 (71%)] | Loss: 0.342162 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:05 | INFO | Rank 0 | Global Steps: 3710/5215 | Train Epoch: 1 [178080/250320 (71%)] | Loss: 0.380985 | Image2Text Acc: 89.58 | Text2Image Acc: 83.33 | Data Time: 0.018s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:08 | INFO | Rank 0 | Global Steps: 3720/5215 | Train Epoch: 1 [178560/250320 (71%)] | Loss: 0.256423 | Image2Text Acc: 89.58 | Text2Image Acc: 93.75 | Data Time: 0.022s | Batch Time: 0.374s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:12 | INFO | Rank 0 | Global Steps: 3730/5215 | Train Epoch: 1 [179040/250320 (72%)] | Loss: 0.194401 | Image2Text Acc: 95.83 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:16 | INFO | Rank 0 | Global Steps: 3740/5215 | Train Epoch: 1 [179520/250320 (72%)] | Loss: 0.087899 | Image2Text Acc: 100.00 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.374s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:19 | INFO | Rank 0 | Global Steps: 3750/5215 | Train Epoch: 1 [180000/250320 (72%)] | Loss: 0.322438 | Image2Text Acc: 91.67 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:23 | INFO | Rank 0 | Global Steps: 3760/5215 | Train Epoch: 1 [180480/250320 (72%)] | Loss: 0.530988 | Image2Text Acc: 87.50 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:27 | INFO | Rank 0 | Global Steps: 3770/5215 | Train Epoch: 1 [180960/250320 (72%)] | Loss: 0.212841 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.380s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:30 | INFO | Rank 0 | Global Steps: 3780/5215 | Train Epoch: 1 [181440/250320 (72%)] | Loss: 0.546179 | Image2Text Acc: 91.67 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:34 | INFO | Rank 0 | Global Steps: 3790/5215 | Train Epoch: 1 [181920/250320 (73%)] | Loss: 0.480567 | Image2Text Acc: 81.25 | Text2Image Acc: 81.25 | Data Time: 0.022s | Batch Time: 0.371s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:38 | INFO | Rank 0 | Global Steps: 3800/5215 | Train Epoch: 1 [182400/250320 (73%)] | Loss: 0.340113 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:41 | INFO | Rank 0 | Global Steps: 3810/5215 | Train Epoch: 1 [182880/250320 (73%)] | Loss: 0.432899 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:45 | INFO | Rank 0 | Global Steps: 3820/5215 | Train Epoch: 1 [183360/250320 (73%)] | Loss: 0.267409 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:49 | INFO | Rank 0 | Global Steps: 3830/5215 | Train Epoch: 1 [183840/250320 (73%)] | Loss: 0.552540 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:52 | INFO | Rank 0 | Global Steps: 3840/5215 | Train Epoch: 1 [184320/250320 (74%)] | Loss: 0.200446 | Image2Text Acc: 95.83 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000001 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:56 | INFO | Rank 0 | Global Steps: 3850/5215 | Train Epoch: 1 [184800/250320 (74%)] | Loss: 0.244700 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:02:59 | INFO | Rank 0 | Global Steps: 3860/5215 | Train Epoch: 1 [185280/250320 (74%)] | Loss: 0.191211 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.022s | Batch Time: 0.382s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:03 | INFO | Rank 0 | Global Steps: 3870/5215 | Train Epoch: 1 [185760/250320 (74%)] | Loss: 0.579678 | Image2Text Acc: 91.67 | Text2Image Acc: 81.25 | Data Time: 0.020s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:07 | INFO | Rank 0 | Global Steps: 3880/5215 | Train Epoch: 1 [186240/250320 (74%)] | Loss: 0.406712 | Image2Text Acc: 93.75 | Text2Image Acc: 87.50 | Data Time: 0.022s | Batch Time: 0.374s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:11 | INFO | Rank 0 | Global Steps: 3890/5215 | Train Epoch: 1 [186720/250320 (75%)] | Loss: 0.292472 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.366s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:14 | INFO | Rank 0 | Global Steps: 3900/5215 | Train Epoch: 1 [187200/250320 (75%)] | Loss: 0.610871 | Image2Text Acc: 79.17 | Text2Image Acc: 79.17 | Data Time: 0.020s | Batch Time: 0.373s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:18 | INFO | Rank 0 | Global Steps: 3910/5215 | Train Epoch: 1 [187680/250320 (75%)] | Loss: 0.273239 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:22 | INFO | Rank 0 | Global Steps: 3920/5215 | Train Epoch: 1 [188160/250320 (75%)] | Loss: 0.226585 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:25 | INFO | Rank 0 | Global Steps: 3930/5215 | Train Epoch: 1 [188640/250320 (75%)] | Loss: 0.306814 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:29 | INFO | Rank 0 | Global Steps: 3940/5215 | Train Epoch: 1 [189120/250320 (76%)] | Loss: 0.193439 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:33 | INFO | Rank 0 | Global Steps: 3950/5215 | Train Epoch: 1 [189600/250320 (76%)] | Loss: 0.365571 | Image2Text Acc: 81.25 | Text2Image Acc: 91.67 | Data Time: 0.021s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:36 | INFO | Rank 0 | Global Steps: 3960/5215 | Train Epoch: 1 [190080/250320 (76%)] | Loss: 0.273761 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.022s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:40 | INFO | Rank 0 | Global Steps: 3970/5215 | Train Epoch: 1 [190560/250320 (76%)] | Loss: 0.455837 | Image2Text Acc: 81.25 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:44 | INFO | Rank 0 | Global Steps: 3980/5215 | Train Epoch: 1 [191040/250320 (76%)] | Loss: 0.410648 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:47 | INFO | Rank 0 | Global Steps: 3990/5215 | Train Epoch: 1 [191520/250320 (77%)] | Loss: 0.544485 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.023s | Batch Time: 0.373s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:51 | INFO | Rank 0 | Global Steps: 4000/5215 | Train Epoch: 1 [192000/250320 (77%)] | Loss: 0.207225 | Image2Text Acc: 93.75 | Text2Image Acc: 95.83 | Data Time: 0.018s | Batch Time: 0.369s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:03:51 | INFO | Rank 0 | Begin to eval on validation set (epoch 1 @ 4000 steps)...\n", + "2023-01-14,21:04:14 | INFO | Rank 0 | Evaluated 100/638 batches...\n", + "2023-01-14,21:04:36 | INFO | Rank 0 | Evaluated 200/638 batches...\n", + "2023-01-14,21:04:59 | INFO | Rank 0 | Evaluated 300/638 batches...\n", + "2023-01-14,21:05:21 | INFO | Rank 0 | Evaluated 400/638 batches...\n", + "2023-01-14,21:05:44 | INFO | Rank 0 | Evaluated 500/638 batches...\n", + "2023-01-14,21:06:06 | INFO | Rank 0 | Evaluated 600/638 batches...\n", + "2023-01-14,21:06:15 | INFO | Rank 0 | Validation Result (epoch 1 @ 4000 steps) | Valid Loss: 0.240804 | Image2Text Acc: 92.66 | Text2Image Acc: 91.94 | logit_scale: 4.605 | Valid Batch Size: 48\n", + "2023-01-14,21:06:18 | INFO | Rank 0 | Global Steps: 4010/5215 | Train Epoch: 1 [192480/250320 (77%)] | Loss: 0.328553 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:22 | INFO | Rank 0 | Global Steps: 4020/5215 | Train Epoch: 1 [192960/250320 (77%)] | Loss: 0.152823 | Image2Text Acc: 95.83 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:26 | INFO | Rank 0 | Global Steps: 4030/5215 | Train Epoch: 1 [193440/250320 (77%)] | Loss: 0.260934 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.021s | Batch Time: 0.371s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:29 | INFO | Rank 0 | Global Steps: 4040/5215 | Train Epoch: 1 [193920/250320 (77%)] | Loss: 0.372008 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.022s | Batch Time: 0.366s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:33 | INFO | Rank 0 | Global Steps: 4050/5215 | Train Epoch: 1 [194400/250320 (78%)] | Loss: 0.442053 | Image2Text Acc: 91.67 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.366s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:37 | INFO | Rank 0 | Global Steps: 4060/5215 | Train Epoch: 1 [194880/250320 (78%)] | Loss: 0.370239 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:40 | INFO | Rank 0 | Global Steps: 4070/5215 | Train Epoch: 1 [195360/250320 (78%)] | Loss: 0.259128 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:44 | INFO | Rank 0 | Global Steps: 4080/5215 | Train Epoch: 1 [195840/250320 (78%)] | Loss: 0.235972 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.021s | Batch Time: 0.379s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:48 | INFO | Rank 0 | Global Steps: 4090/5215 | Train Epoch: 1 [196320/250320 (78%)] | Loss: 0.232661 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:51 | INFO | Rank 0 | Global Steps: 4100/5215 | Train Epoch: 1 [196800/250320 (79%)] | Loss: 0.323244 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.021s | Batch Time: 0.367s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:55 | INFO | Rank 0 | Global Steps: 4110/5215 | Train Epoch: 1 [197280/250320 (79%)] | Loss: 0.220095 | Image2Text Acc: 95.83 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:06:59 | INFO | Rank 0 | Global Steps: 4120/5215 | Train Epoch: 1 [197760/250320 (79%)] | Loss: 0.243769 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:02 | INFO | Rank 0 | Global Steps: 4130/5215 | Train Epoch: 1 [198240/250320 (79%)] | Loss: 0.458951 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.024s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:06 | INFO | Rank 0 | Global Steps: 4140/5215 | Train Epoch: 1 [198720/250320 (79%)] | Loss: 0.298845 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.021s | Batch Time: 0.375s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:10 | INFO | Rank 0 | Global Steps: 4150/5215 | Train Epoch: 1 [199200/250320 (80%)] | Loss: 0.467961 | Image2Text Acc: 91.67 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:13 | INFO | Rank 0 | Global Steps: 4160/5215 | Train Epoch: 1 [199680/250320 (80%)] | Loss: 0.436175 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:17 | INFO | Rank 0 | Global Steps: 4170/5215 | Train Epoch: 1 [200160/250320 (80%)] | Loss: 0.262451 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:21 | INFO | Rank 0 | Global Steps: 4180/5215 | Train Epoch: 1 [200640/250320 (80%)] | Loss: 0.545027 | Image2Text Acc: 85.42 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:24 | INFO | Rank 0 | Global Steps: 4190/5215 | Train Epoch: 1 [201120/250320 (80%)] | Loss: 0.564339 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:28 | INFO | Rank 0 | Global Steps: 4200/5215 | Train Epoch: 1 [201600/250320 (81%)] | Loss: 0.627127 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:32 | INFO | Rank 0 | Global Steps: 4210/5215 | Train Epoch: 1 [202080/250320 (81%)] | Loss: 0.324819 | Image2Text Acc: 95.83 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:35 | INFO | Rank 0 | Global Steps: 4220/5215 | Train Epoch: 1 [202560/250320 (81%)] | Loss: 0.407902 | Image2Text Acc: 83.33 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:39 | INFO | Rank 0 | Global Steps: 4230/5215 | Train Epoch: 1 [203040/250320 (81%)] | Loss: 0.320982 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.023s | Batch Time: 0.392s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:43 | INFO | Rank 0 | Global Steps: 4240/5215 | Train Epoch: 1 [203520/250320 (81%)] | Loss: 0.484319 | Image2Text Acc: 85.42 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.368s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:46 | INFO | Rank 0 | Global Steps: 4250/5215 | Train Epoch: 1 [204000/250320 (81%)] | Loss: 0.147294 | Image2Text Acc: 95.83 | Text2Image Acc: 100.00 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:50 | INFO | Rank 0 | Global Steps: 4260/5215 | Train Epoch: 1 [204480/250320 (82%)] | Loss: 0.169659 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.375s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:54 | INFO | Rank 0 | Global Steps: 4270/5215 | Train Epoch: 1 [204960/250320 (82%)] | Loss: 0.272106 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.023s | Batch Time: 0.377s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:07:57 | INFO | Rank 0 | Global Steps: 4280/5215 | Train Epoch: 1 [205440/250320 (82%)] | Loss: 0.326259 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:01 | INFO | Rank 0 | Global Steps: 4290/5215 | Train Epoch: 1 [205920/250320 (82%)] | Loss: 0.337684 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:05 | INFO | Rank 0 | Global Steps: 4300/5215 | Train Epoch: 1 [206400/250320 (82%)] | Loss: 0.381485 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:08 | INFO | Rank 0 | Global Steps: 4310/5215 | Train Epoch: 1 [206880/250320 (83%)] | Loss: 0.256099 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:12 | INFO | Rank 0 | Global Steps: 4320/5215 | Train Epoch: 1 [207360/250320 (83%)] | Loss: 0.373579 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:16 | INFO | Rank 0 | Global Steps: 4330/5215 | Train Epoch: 1 [207840/250320 (83%)] | Loss: 0.245633 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:19 | INFO | Rank 0 | Global Steps: 4340/5215 | Train Epoch: 1 [208320/250320 (83%)] | Loss: 0.499349 | Image2Text Acc: 83.33 | Text2Image Acc: 85.42 | Data Time: 0.026s | Batch Time: 0.368s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:23 | INFO | Rank 0 | Global Steps: 4350/5215 | Train Epoch: 1 [208800/250320 (83%)] | Loss: 0.190400 | Image2Text Acc: 93.75 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:27 | INFO | Rank 0 | Global Steps: 4360/5215 | Train Epoch: 1 [209280/250320 (84%)] | Loss: 0.299177 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:30 | INFO | Rank 0 | Global Steps: 4370/5215 | Train Epoch: 1 [209760/250320 (84%)] | Loss: 0.320892 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:34 | INFO | Rank 0 | Global Steps: 4380/5215 | Train Epoch: 1 [210240/250320 (84%)] | Loss: 0.260111 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:38 | INFO | Rank 0 | Global Steps: 4390/5215 | Train Epoch: 1 [210720/250320 (84%)] | Loss: 0.251131 | Image2Text Acc: 95.83 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:41 | INFO | Rank 0 | Global Steps: 4400/5215 | Train Epoch: 1 [211200/250320 (84%)] | Loss: 0.288052 | Image2Text Acc: 87.50 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:45 | INFO | Rank 0 | Global Steps: 4410/5215 | Train Epoch: 1 [211680/250320 (85%)] | Loss: 0.236768 | Image2Text Acc: 91.67 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:48 | INFO | Rank 0 | Global Steps: 4420/5215 | Train Epoch: 1 [212160/250320 (85%)] | Loss: 0.592548 | Image2Text Acc: 79.17 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:52 | INFO | Rank 0 | Global Steps: 4430/5215 | Train Epoch: 1 [212640/250320 (85%)] | Loss: 0.320176 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.367s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:08:56 | INFO | Rank 0 | Global Steps: 4440/5215 | Train Epoch: 1 [213120/250320 (85%)] | Loss: 0.205553 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.022s | Batch Time: 0.366s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:00 | INFO | Rank 0 | Global Steps: 4450/5215 | Train Epoch: 1 [213600/250320 (85%)] | Loss: 0.316974 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.030s | Batch Time: 0.372s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:03 | INFO | Rank 0 | Global Steps: 4460/5215 | Train Epoch: 1 [214080/250320 (86%)] | Loss: 0.160544 | Image2Text Acc: 93.75 | Text2Image Acc: 97.92 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:07 | INFO | Rank 0 | Global Steps: 4470/5215 | Train Epoch: 1 [214560/250320 (86%)] | Loss: 0.241178 | Image2Text Acc: 91.67 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:11 | INFO | Rank 0 | Global Steps: 4480/5215 | Train Epoch: 1 [215040/250320 (86%)] | Loss: 0.325726 | Image2Text Acc: 93.75 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.373s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:14 | INFO | Rank 0 | Global Steps: 4490/5215 | Train Epoch: 1 [215520/250320 (86%)] | Loss: 0.446931 | Image2Text Acc: 85.42 | Text2Image Acc: 87.50 | Data Time: 0.023s | Batch Time: 0.376s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:18 | INFO | Rank 0 | Global Steps: 4500/5215 | Train Epoch: 1 [216000/250320 (86%)] | Loss: 0.515944 | Image2Text Acc: 89.58 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:22 | INFO | Rank 0 | Global Steps: 4510/5215 | Train Epoch: 1 [216480/250320 (86%)] | Loss: 0.436900 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:25 | INFO | Rank 0 | Global Steps: 4520/5215 | Train Epoch: 1 [216960/250320 (87%)] | Loss: 0.395753 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:29 | INFO | Rank 0 | Global Steps: 4530/5215 | Train Epoch: 1 [217440/250320 (87%)] | Loss: 0.298280 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.037s | Batch Time: 0.394s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:32 | INFO | Rank 0 | Global Steps: 4540/5215 | Train Epoch: 1 [217920/250320 (87%)] | Loss: 0.297875 | Image2Text Acc: 93.75 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:36 | INFO | Rank 0 | Global Steps: 4550/5215 | Train Epoch: 1 [218400/250320 (87%)] | Loss: 0.368746 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:40 | INFO | Rank 0 | Global Steps: 4560/5215 | Train Epoch: 1 [218880/250320 (87%)] | Loss: 0.299198 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:43 | INFO | Rank 0 | Global Steps: 4570/5215 | Train Epoch: 1 [219360/250320 (88%)] | Loss: 0.391412 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.022s | Batch Time: 0.371s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:47 | INFO | Rank 0 | Global Steps: 4580/5215 | Train Epoch: 1 [219840/250320 (88%)] | Loss: 0.253173 | Image2Text Acc: 95.83 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:51 | INFO | Rank 0 | Global Steps: 4590/5215 | Train Epoch: 1 [220320/250320 (88%)] | Loss: 0.298226 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:54 | INFO | Rank 0 | Global Steps: 4600/5215 | Train Epoch: 1 [220800/250320 (88%)] | Loss: 0.505762 | Image2Text Acc: 81.25 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:09:58 | INFO | Rank 0 | Global Steps: 4610/5215 | Train Epoch: 1 [221280/250320 (88%)] | Loss: 0.368106 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:02 | INFO | Rank 0 | Global Steps: 4620/5215 | Train Epoch: 1 [221760/250320 (89%)] | Loss: 0.099810 | Image2Text Acc: 95.83 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:05 | INFO | Rank 0 | Global Steps: 4630/5215 | Train Epoch: 1 [222240/250320 (89%)] | Loss: 0.271291 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:09 | INFO | Rank 0 | Global Steps: 4640/5215 | Train Epoch: 1 [222720/250320 (89%)] | Loss: 0.241306 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:13 | INFO | Rank 0 | Global Steps: 4650/5215 | Train Epoch: 1 [223200/250320 (89%)] | Loss: 0.393220 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:16 | INFO | Rank 0 | Global Steps: 4660/5215 | Train Epoch: 1 [223680/250320 (89%)] | Loss: 0.196638 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:20 | INFO | Rank 0 | Global Steps: 4670/5215 | Train Epoch: 1 [224160/250320 (90%)] | Loss: 0.207612 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:23 | INFO | Rank 0 | Global Steps: 4680/5215 | Train Epoch: 1 [224640/250320 (90%)] | Loss: 0.534915 | Image2Text Acc: 79.17 | Text2Image Acc: 83.33 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:27 | INFO | Rank 0 | Global Steps: 4690/5215 | Train Epoch: 1 [225120/250320 (90%)] | Loss: 0.335680 | Image2Text Acc: 95.83 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:31 | INFO | Rank 0 | Global Steps: 4700/5215 | Train Epoch: 1 [225600/250320 (90%)] | Loss: 0.379060 | Image2Text Acc: 83.33 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:34 | INFO | Rank 0 | Global Steps: 4710/5215 | Train Epoch: 1 [226080/250320 (90%)] | Loss: 0.195548 | Image2Text Acc: 91.67 | Text2Image Acc: 95.83 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:38 | INFO | Rank 0 | Global Steps: 4720/5215 | Train Epoch: 1 [226560/250320 (91%)] | Loss: 0.206264 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:42 | INFO | Rank 0 | Global Steps: 4730/5215 | Train Epoch: 1 [227040/250320 (91%)] | Loss: 0.344749 | Image2Text Acc: 89.58 | Text2Image Acc: 85.42 | Data Time: 0.020s | Batch Time: 0.370s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:45 | INFO | Rank 0 | Global Steps: 4740/5215 | Train Epoch: 1 [227520/250320 (91%)] | Loss: 0.401616 | Image2Text Acc: 93.75 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:49 | INFO | Rank 0 | Global Steps: 4750/5215 | Train Epoch: 1 [228000/250320 (91%)] | Loss: 0.193034 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.385s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:53 | INFO | Rank 0 | Global Steps: 4760/5215 | Train Epoch: 1 [228480/250320 (91%)] | Loss: 0.269970 | Image2Text Acc: 93.75 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:10:56 | INFO | Rank 0 | Global Steps: 4770/5215 | Train Epoch: 1 [228960/250320 (91%)] | Loss: 0.498633 | Image2Text Acc: 85.42 | Text2Image Acc: 81.25 | Data Time: 0.021s | Batch Time: 0.370s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:00 | INFO | Rank 0 | Global Steps: 4780/5215 | Train Epoch: 1 [229440/250320 (92%)] | Loss: 0.451746 | Image2Text Acc: 81.25 | Text2Image Acc: 85.42 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:04 | INFO | Rank 0 | Global Steps: 4790/5215 | Train Epoch: 1 [229920/250320 (92%)] | Loss: 0.455822 | Image2Text Acc: 81.25 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:07 | INFO | Rank 0 | Global Steps: 4800/5215 | Train Epoch: 1 [230400/250320 (92%)] | Loss: 0.270321 | Image2Text Acc: 91.67 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:11 | INFO | Rank 0 | Global Steps: 4810/5215 | Train Epoch: 1 [230880/250320 (92%)] | Loss: 0.382001 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:15 | INFO | Rank 0 | Global Steps: 4820/5215 | Train Epoch: 1 [231360/250320 (92%)] | Loss: 0.338040 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.021s | Batch Time: 0.370s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:18 | INFO | Rank 0 | Global Steps: 4830/5215 | Train Epoch: 1 [231840/250320 (93%)] | Loss: 0.218796 | Image2Text Acc: 89.58 | Text2Image Acc: 95.83 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:22 | INFO | Rank 0 | Global Steps: 4840/5215 | Train Epoch: 1 [232320/250320 (93%)] | Loss: 0.211760 | Image2Text Acc: 95.83 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:26 | INFO | Rank 0 | Global Steps: 4850/5215 | Train Epoch: 1 [232800/250320 (93%)] | Loss: 0.231990 | Image2Text Acc: 91.67 | Text2Image Acc: 95.83 | Data Time: 0.019s | Batch Time: 0.378s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:29 | INFO | Rank 0 | Global Steps: 4860/5215 | Train Epoch: 1 [233280/250320 (93%)] | Loss: 0.341754 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:33 | INFO | Rank 0 | Global Steps: 4870/5215 | Train Epoch: 1 [233760/250320 (93%)] | Loss: 0.143991 | Image2Text Acc: 93.75 | Text2Image Acc: 97.92 | Data Time: 0.018s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:37 | INFO | Rank 0 | Global Steps: 4880/5215 | Train Epoch: 1 [234240/250320 (94%)] | Loss: 0.280634 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:40 | INFO | Rank 0 | Global Steps: 4890/5215 | Train Epoch: 1 [234720/250320 (94%)] | Loss: 0.405270 | Image2Text Acc: 85.42 | Text2Image Acc: 93.75 | Data Time: 0.021s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:44 | INFO | Rank 0 | Global Steps: 4900/5215 | Train Epoch: 1 [235200/250320 (94%)] | Loss: 0.637904 | Image2Text Acc: 87.50 | Text2Image Acc: 79.17 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:48 | INFO | Rank 0 | Global Steps: 4910/5215 | Train Epoch: 1 [235680/250320 (94%)] | Loss: 0.244760 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:51 | INFO | Rank 0 | Global Steps: 4920/5215 | Train Epoch: 1 [236160/250320 (94%)] | Loss: 0.276766 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.365s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:55 | INFO | Rank 0 | Global Steps: 4930/5215 | Train Epoch: 1 [236640/250320 (95%)] | Loss: 0.377489 | Image2Text Acc: 89.58 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:11:59 | INFO | Rank 0 | Global Steps: 4940/5215 | Train Epoch: 1 [237120/250320 (95%)] | Loss: 0.329878 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.374s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:12:02 | INFO | Rank 0 | Global Steps: 4950/5215 | Train Epoch: 1 [237600/250320 (95%)] | Loss: 0.462497 | Image2Text Acc: 83.33 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:12:06 | INFO | Rank 0 | Global Steps: 4960/5215 | Train Epoch: 1 [238080/250320 (95%)] | Loss: 0.281811 | Image2Text Acc: 93.75 | Text2Image Acc: 85.42 | Data Time: 0.018s | Batch Time: 0.360s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:12:10 | INFO | Rank 0 | Global Steps: 4970/5215 | Train Epoch: 1 [238560/250320 (95%)] | Loss: 0.368849 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:12:13 | INFO | Rank 0 | Global Steps: 4980/5215 | Train Epoch: 1 [239040/250320 (95%)] | Loss: 0.242054 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.370s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:12:17 | INFO | Rank 0 | Global Steps: 4990/5215 | Train Epoch: 1 [239520/250320 (96%)] | Loss: 0.392416 | Image2Text Acc: 87.50 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:12:21 | INFO | Rank 0 | Global Steps: 5000/5215 | Train Epoch: 1 [240000/250320 (96%)] | Loss: 0.260949 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:12:21 | INFO | Rank 0 | Begin to eval on validation set (epoch 1 @ 5000 steps)...\n", + "2023-01-14,21:12:44 | INFO | Rank 0 | Evaluated 100/638 batches...\n", + "2023-01-14,21:13:06 | INFO | Rank 0 | Evaluated 200/638 batches...\n", + "2023-01-14,21:13:28 | INFO | Rank 0 | Evaluated 300/638 batches...\n", + "2023-01-14,21:13:51 | INFO | Rank 0 | Evaluated 400/638 batches...\n", + "2023-01-14,21:14:13 | INFO | Rank 0 | Evaluated 500/638 batches...\n", + "2023-01-14,21:14:36 | INFO | Rank 0 | Evaluated 600/638 batches...\n", + "2023-01-14,21:14:44 | INFO | Rank 0 | Validation Result (epoch 1 @ 5000 steps) | Valid Loss: 0.238828 | Image2Text Acc: 92.71 | Text2Image Acc: 92.00 | logit_scale: 4.605 | Valid Batch Size: 48\n", + "2023-01-14,21:14:48 | INFO | Rank 0 | Global Steps: 5010/5215 | Train Epoch: 1 [240480/250320 (96%)] | Loss: 0.296646 | Image2Text Acc: 91.67 | Text2Image Acc: 93.75 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:14:52 | INFO | Rank 0 | Global Steps: 5020/5215 | Train Epoch: 1 [240960/250320 (96%)] | Loss: 0.212507 | Image2Text Acc: 87.50 | Text2Image Acc: 93.75 | Data Time: 0.031s | Batch Time: 0.385s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:14:56 | INFO | Rank 0 | Global Steps: 5030/5215 | Train Epoch: 1 [241440/250320 (96%)] | Loss: 0.187542 | Image2Text Acc: 93.75 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:14:59 | INFO | Rank 0 | Global Steps: 5040/5215 | Train Epoch: 1 [241920/250320 (97%)] | Loss: 0.174362 | Image2Text Acc: 95.83 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:03 | INFO | Rank 0 | Global Steps: 5050/5215 | Train Epoch: 1 [242400/250320 (97%)] | Loss: 0.396933 | Image2Text Acc: 83.33 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:07 | INFO | Rank 0 | Global Steps: 5060/5215 | Train Epoch: 1 [242880/250320 (97%)] | Loss: 0.190303 | Image2Text Acc: 95.83 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:10 | INFO | Rank 0 | Global Steps: 5070/5215 | Train Epoch: 1 [243360/250320 (97%)] | Loss: 0.200660 | Image2Text Acc: 93.75 | Text2Image Acc: 95.83 | Data Time: 0.026s | Batch Time: 0.369s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:14 | INFO | Rank 0 | Global Steps: 5080/5215 | Train Epoch: 1 [243840/250320 (97%)] | Loss: 0.560694 | Image2Text Acc: 87.50 | Text2Image Acc: 83.33 | Data Time: 0.019s | Batch Time: 0.361s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:18 | INFO | Rank 0 | Global Steps: 5090/5215 | Train Epoch: 1 [244320/250320 (98%)] | Loss: 0.354665 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:21 | INFO | Rank 0 | Global Steps: 5100/5215 | Train Epoch: 1 [244800/250320 (98%)] | Loss: 0.173221 | Image2Text Acc: 97.92 | Text2Image Acc: 93.75 | Data Time: 0.019s | Batch Time: 0.364s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:25 | INFO | Rank 0 | Global Steps: 5110/5215 | Train Epoch: 1 [245280/250320 (98%)] | Loss: 0.559426 | Image2Text Acc: 81.25 | Text2Image Acc: 79.17 | Data Time: 0.021s | Batch Time: 0.371s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:28 | INFO | Rank 0 | Global Steps: 5120/5215 | Train Epoch: 1 [245760/250320 (98%)] | Loss: 0.356126 | Image2Text Acc: 89.58 | Text2Image Acc: 89.58 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:32 | INFO | Rank 0 | Global Steps: 5130/5215 | Train Epoch: 1 [246240/250320 (98%)] | Loss: 0.320622 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.022s | Batch Time: 0.364s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:36 | INFO | Rank 0 | Global Steps: 5140/5215 | Train Epoch: 1 [246720/250320 (99%)] | Loss: 0.293673 | Image2Text Acc: 91.67 | Text2Image Acc: 89.58 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:39 | INFO | Rank 0 | Global Steps: 5150/5215 | Train Epoch: 1 [247200/250320 (99%)] | Loss: 0.752282 | Image2Text Acc: 83.33 | Text2Image Acc: 81.25 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:43 | INFO | Rank 0 | Global Steps: 5160/5215 | Train Epoch: 1 [247680/250320 (99%)] | Loss: 0.363567 | Image2Text Acc: 87.50 | Text2Image Acc: 87.50 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:47 | INFO | Rank 0 | Global Steps: 5170/5215 | Train Epoch: 1 [248160/250320 (99%)] | Loss: 0.731255 | Image2Text Acc: 83.33 | Text2Image Acc: 77.08 | Data Time: 0.020s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:50 | INFO | Rank 0 | Global Steps: 5180/5215 | Train Epoch: 1 [248640/250320 (99%)] | Loss: 0.852196 | Image2Text Acc: 70.83 | Text2Image Acc: 72.92 | Data Time: 0.019s | Batch Time: 0.363s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:54 | INFO | Rank 0 | Global Steps: 5190/5215 | Train Epoch: 1 [249120/250320 (100%)] | Loss: 0.407889 | Image2Text Acc: 87.50 | Text2Image Acc: 91.67 | Data Time: 0.019s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:15:58 | INFO | Rank 0 | Global Steps: 5200/5215 | Train Epoch: 1 [249600/250320 (100%)] | Loss: 0.440937 | Image2Text Acc: 89.58 | Text2Image Acc: 87.50 | Data Time: 0.020s | Batch Time: 0.362s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:16:01 | INFO | Rank 0 | Global Steps: 5210/5215 | Train Epoch: 1 [250080/250320 (100%)] | Loss: 0.286732 | Image2Text Acc: 93.75 | Text2Image Acc: 89.58 | Data Time: 0.021s | Batch Time: 0.367s | LR: 0.000000 | logit_scale: 4.605 | Global Batch Size: 48\n", + "2023-01-14,21:16:03 | INFO | Rank 0 | Begin to eval on validation set (epoch 1 @ 5215 steps)...\n", + "2023-01-14,21:16:27 | INFO | Rank 0 | Evaluated 100/638 batches...\n", + "2023-01-14,21:16:49 | INFO | Rank 0 | Evaluated 200/638 batches...\n", + "2023-01-14,21:17:11 | INFO | Rank 0 | Evaluated 300/638 batches...\n", + "2023-01-14,21:17:34 | INFO | Rank 0 | Evaluated 400/638 batches...\n", + "2023-01-14,21:17:56 | INFO | Rank 0 | Evaluated 500/638 batches...\n", + "2023-01-14,21:18:19 | INFO | Rank 0 | Evaluated 600/638 batches...\n", + "2023-01-14,21:18:27 | INFO | Rank 0 | Validation Result (epoch 1 @ 5215 steps) | Valid Loss: 0.238827 | Image2Text Acc: 92.72 | Text2Image Acc: 92.00 | logit_scale: 4.605 | Valid Batch Size: 48\n", + "2023-01-14,21:18:33 | INFO | Rank 0 | Saved checkpoint ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch1.pt (epoch 1 @ 5215 steps) (writing took 5.584172010421753 seconds)\n", + "2023-01-14,21:18:39 | INFO | Rank 0 | Saved checkpoint ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt (epoch 1 @ 5215 steps) (writing took 5.848175048828125 seconds)\n", + "Exception in thread Thread-1:\n", + "Traceback (most recent call last):\n", + " File \"/home/pai/lib/python3.6/threading.py\", line 916, in _bootstrap_inner\n", + " self.run()\n", + " File \"/home/pai/lib/python3.6/threading.py\", line 864, in run\n", + " self._target(*self._args, **self._kwargs)\n", + " File \"/home/pai/lib/python3.6/logging/handlers.py\", line 1476, in _monitor\n", + " record = self.dequeue(True)\n", + " File \"/home/pai/lib/python3.6/logging/handlers.py\", line 1425, in dequeue\n", + " return self.queue.get(block)\n", + " File \"/home/pai/lib/python3.6/multiprocessing/queues.py\", line 94, in get\n", + " res = self._recv_bytes()\n", + " File \"/home/pai/lib/python3.6/multiprocessing/connection.py\", line 216, in recv_bytes\n", + " buf = self._recv_bytes(maxlength)\n", + " File \"/home/pai/lib/python3.6/multiprocessing/connection.py\", line 407, in _recv_bytes\n", + " buf = self._recv(4)\n", + " File \"/home/pai/lib/python3.6/multiprocessing/connection.py\", line 383, in _recv\n", + " raise EOFError\n", + "EOFError\n", + "\n" + ] + } + ], + "source": [ + "# 执行finetune流程\n", + "!{run_command}" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 验证集效果验证\n", + "\n", + "为了验证模型的效果，我们提供特征提取、以及图文检索任务评估的流程。更加详尽的描述，也可参考Github readme的相关部分说明：https://github.com/OFA-Sys/Chinese-CLIP/blob/master/README.md#预测及评估。\n", + "\n", + "### 1. 计算图文特征" + ] + }, + { + "cell_type": "code", + "execution_count": 39, + "metadata": { + "ExecutionIndicator": { + "show": true + }, + "execution": { + "iopub.execute_input": "2023-01-14T13:18:50.103797Z", + "iopub.status.busy": "2023-01-14T13:18:50.103514Z", + "iopub.status.idle": "2023-01-14T13:22:17.287273Z", + "shell.execute_reply": "2023-01-14T13:22:17.286358Z", + "shell.execute_reply.started": "2023-01-14T13:18:50.103769Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\n", + "python -u cn_clip/eval/extract_features.py --extract-image-feats --extract-text-feats --image-data=\"../datapath//datasets/MUGE/lmdb/valid/imgs\" --text-data=\"../datapath//datasets/MUGE/valid_texts.jsonl\" --img-batch-size=32 --text-batch-size=32 --context-length=52 --resume=../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt --vision-model=ViT-B-16 --text-model=RoBERTa-wwm-ext-base-chinese\n", + "\n", + "/home/pai/lib/python3.6/site-packages/OpenSSL/crypto.py:12: CryptographyDeprecationWarning: Python 3.6 is no longer supported by the Python core team. Therefore, support for it is deprecated in cryptography and will be removed in a future release.\n", + " from cryptography import x509\n", + "Params:\n", + " context_length: 52\n", + " debug: False\n", + " extract_image_feats: True\n", + " extract_text_feats: True\n", + " image_data: ../datapath//datasets/MUGE/lmdb/valid/imgs\n", + " image_feat_output_path: None\n", + " img_batch_size: 32\n", + " precision: amp\n", + " resume: ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt\n", + " text_batch_size: 32\n", + " text_data: ../datapath//datasets/MUGE/valid_texts.jsonl\n", + " text_feat_output_path: None\n", + " text_model: RoBERTa-wwm-ext-base-chinese\n", + " vision_model: ViT-B-16\n", + "Loading vision model config from cn_clip/clip/model_configs/ViT-B-16.json\n", + "Loading text model config from cn_clip/clip/model_configs/RoBERTa-wwm-ext-base-chinese.json\n", + "Preparing image inference dataset.\n", + "Preparing text inference dataset.\n", + "Begin to load model checkpoint from ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt.\n", + "=> loaded checkpoint '../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt' (epoch 1 @ 5215 steps)\n", + "Make inference for texts...\n", + "100%|█████████████████████████████████████████| 157/157 [00:07<00:00, 21.66it/s]\n", + "5008 text features are stored in ../datapath//datasets/MUGE/valid_texts.txt_feat.jsonl\n", + "Make inference for images...\n", + "100%|█████████████████████████████████████████| 932/932 [02:56<00:00, 5.28it/s]\n", + "29806 image features are stored in ../datapath//datasets/MUGE/valid_imgs.img_feat.jsonl\n", + "Done!\n" + ] + } + ], + "source": [ + "# 为验证集图片池和query文本计算特征\n", + "dataset_name=\"MUGE\"\n", + "split=\"valid\" # 指定计算valid或test集特征\n", + "# 指定我们刚刚finetune的ckpt路径，也可以指定预训练ckpt路径测试zero-shot效果\n", + "resume=f\"{DATAPATH}/experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt\"\n", + "\n", + "run_command = \"export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\" + \\\n", + "f\"\"\"\n", + "python -u cn_clip/eval/extract_features.py \\\n", + " --extract-image-feats \\\n", + " --extract-text-feats \\\n", + " --image-data=\"{DATAPATH}/datasets/{dataset_name}/lmdb/{split}/imgs\" \\\n", + " --text-data=\"{DATAPATH}/datasets/{dataset_name}/{split}_texts.jsonl\" \\\n", + " --img-batch-size=32 \\\n", + " --text-batch-size=32 \\\n", + " --context-length=52 \\\n", + " --resume={resume} \\\n", + " --vision-model=ViT-B-16 \\\n", + " --text-model=RoBERTa-wwm-ext-base-chinese\n", + "\"\"\"\n", + "print(run_command.lstrip())\n", + "!{run_command}" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "产出图文特征将保存于`../datapath/datasets/MUGE`目录下，图片特征保存于`valid_imgs.img_feat.jsonl`文件，每行以json存储一张图片的特征，格式如下：" + ] + }, + { + "cell_type": "raw", + "metadata": { + "ExecutionIndicator": { + "show": false + }, + "execution": { + "iopub.execute_input": "2023-01-14T13:23:11.828610Z", + "iopub.status.busy": "2023-01-14T13:23:11.828327Z", + "iopub.status.idle": "2023-01-14T13:23:12.128343Z", + "shell.execute_reply": "2023-01-14T13:23:12.127477Z", + "shell.execute_reply.started": "2023-01-14T13:23:11.828584Z" + } + }, + "source": [ + "{\"image_id\": 1000002, \"feature\": [0.0198, ..., -0.017, 0.0248]}" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "文本特征则保存于`valid_texts.txt_feat.jsonl`，格式如下：" + ] + }, + { + "cell_type": "raw", + "metadata": {}, + "source": [ + "{\"text_id\": 248816, \"feature\": [0.1314, ..., 0.0018, -0.0002]}" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### 2. KNN检索\n", + "\n", + "对于小规模的学术检索数据集，我们提供一个简单的KNN检索实现，便于计算文到图检索，在验证集3w图片池的top-k召回结果。" + ] + }, + { + "cell_type": "code", + "execution_count": 43, + "metadata": { + "ExecutionIndicator": { + "show": true + }, + "execution": { + "iopub.execute_input": "2023-01-14T13:24:10.023574Z", + "iopub.status.busy": "2023-01-14T13:24:10.023294Z", + "iopub.status.idle": "2023-01-14T13:27:02.536015Z", + "shell.execute_reply": "2023-01-14T13:27:02.535061Z", + "shell.execute_reply.started": "2023-01-14T13:24:10.023548Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\n", + "python -u cn_clip/eval/make_topk_predictions.py --image-feats=\"../datapath//datasets/MUGE/valid_imgs.img_feat.jsonl\" --text-feats=\"../datapath//datasets/MUGE/valid_texts.txt_feat.jsonl\" --top-k=10 --eval-batch-size=32768 --output=\"../datapath//datasets/MUGE/valid_predictions.jsonl\"\n", + "\n", + "Params:\n", + " eval_batch_size: 32768\n", + " image_feats: ../datapath//datasets/MUGE/valid_imgs.img_feat.jsonl\n", + " output: ../datapath//datasets/MUGE/valid_predictions.jsonl\n", + " text_feats: ../datapath//datasets/MUGE/valid_texts.txt_feat.jsonl\n", + " top_k: 10\n", + "Begin to load image features...\n", + "29806it [00:07, 3768.32it/s]\n", + "Finished loading image features.\n", + "Begin to compute top-10 predictions for texts...\n", + "5008it [02:40, 31.17it/s]\n", + "Top-10 predictions are saved in ../datapath//datasets/MUGE/valid_predictions.jsonl\n", + "Done!\n" + ] + } + ], + "source": [ + "# 进行KNN检索，为验证集每个query，匹配特征余弦相似度最高的top-10商品图片\n", + "\n", + "split=\"valid\" # 指定计算valid或test集特征\n", + "\n", + "run_command = \"export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\" + \\\n", + "f\"\"\"\n", + "python -u cn_clip/eval/make_topk_predictions.py \\\n", + " --image-feats=\"{DATAPATH}/datasets/{dataset_name}/{split}_imgs.img_feat.jsonl\" \\\n", + " --text-feats=\"{DATAPATH}/datasets/{dataset_name}/{split}_texts.txt_feat.jsonl\" \\\n", + " --top-k=10 \\\n", + " --eval-batch-size=32768 \\\n", + " --output=\"{DATAPATH}/datasets/{dataset_name}/{split}_predictions.jsonl\"\n", + "\"\"\"\n", + "print(run_command.lstrip())\n", + "!{run_command}" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "产出的结果保存在指定的jsonl文件中，每行表示一个文本召回的top-k图片id（按模型相关性预测分数由大到小排好了序），格式如下，这个格式和我们比赛要求的提交格式是一样的：" + ] + }, + { + "cell_type": "code", + "execution_count": 44, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-14T13:27:27.226086Z", + "iopub.status.busy": "2023-01-14T13:27:27.225694Z", + "iopub.status.idle": "2023-01-14T13:27:27.528758Z", + "shell.execute_reply": "2023-01-14T13:27:27.527955Z", + "shell.execute_reply.started": "2023-01-14T13:27:27.226043Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "{\"text_id\": 248816, \"image_ids\": [286314, 183846, 141999, 936929, 1006938, 162268, 412925, 384823, 877108, 103269]}\n", + "{\"text_id\": 248859, \"image_ids\": [574241, 175548, 678269, 1059854, 708768, 1003303, 854191, 913653, 495295, 153682]}\n", + "{\"text_id\": 248871, \"image_ids\": [60015, 807177, 160459, 706996, 161417, 666622, 637011, 77996, 690344, 783255]}\n", + "{\"text_id\": 248898, \"image_ids\": [946372, 397154, 642386, 777624, 27450, 829271, 222468, 420283, 323919, 937205]}\n", + "{\"text_id\": 248931, \"image_ids\": [225106, 139500, 349934, 941660, 515959, 818516, 646440, 979324, 94740, 1024375]}\n" + ] + } + ], + "source": [ + "!head -n 5 ../datapath/datasets/MUGE/valid_predictions.jsonl" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "我们具体观察其中一组case的预测结果：" + ] + }, + { + "cell_type": "code", + "execution_count": 48, + "metadata": { + "ExecutionIndicator": { + "show": true + }, + "execution": { + "iopub.execute_input": "2023-01-14T13:30:55.918737Z", + "iopub.status.busy": "2023-01-14T13:30:55.918472Z", + "iopub.status.idle": "2023-01-14T13:30:56.283398Z", + "shell.execute_reply": "2023-01-14T13:30:56.282738Z", + "shell.execute_reply.started": "2023-01-14T13:30:55.918712Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "{\"text_id\": 248931, \"text\": \"32不粘锅\", \"image_ids\": [227953, 349934, 646440, 204288, 941660, 425873]}\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# 查看数据样例\n", + "!sed -n \"5,5p\" ../datapath/datasets/MUGE/valid_texts.jsonl # 验证集第三条样本，对应上面第5行\n", + "\n", + "import lmdb\n", + "import base64\n", + "from io import BytesIO\n", + "from PIL import Image\n", + "\n", + "image_ids = [225106, 139500, 349934, 941660, 515959] # 模型预测的top-5相关图片\n", + "\n", + "lmdb_imgs = \"../datapath/datasets/MUGE/lmdb/valid/imgs\"\n", + "env_imgs = lmdb.open(lmdb_imgs, readonly=True, create=False, lock=False, readahead=False, meminit=False)\n", + "txn_imgs = env_imgs.begin(buffers=True)\n", + "for image_id in image_ids:\n", + " image_b64 = txn_imgs.get(\"{}\".format(image_id).encode('utf-8')).tobytes()\n", + " img = Image.open(BytesIO(base64.urlsafe_b64decode(image_b64)))\n", + " img.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### 3. Recall计算\n", + "\n", + "我们提供了评测脚本计算检索任务的Recall@1/5/10，同时给出mean recall（Recall@1/5/10的平均数）。运行如下命令即可获取分数:" + ] + }, + { + "cell_type": "code", + "execution_count": 49, + "metadata": { + "ExecutionIndicator": { + "show": true + }, + "execution": { + "iopub.execute_input": "2023-01-14T13:32:42.857040Z", + "iopub.status.busy": "2023-01-14T13:32:42.856665Z", + "iopub.status.idle": "2023-01-14T13:32:43.610919Z", + "shell.execute_reply": "2023-01-14T13:32:43.609924Z", + "shell.execute_reply.started": "2023-01-14T13:32:42.857004Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\n", + "python cn_clip/eval/evaluation.py ../datapath//datasets/MUGE/valid_texts.jsonl ../datapath//datasets/MUGE/valid_predictions.jsonl output.json;\n", + "cat output.json\n", + "\n", + "Read standard from ../datapath//datasets/MUGE/valid_texts.jsonl\n", + "Read user submit file from ../datapath//datasets/MUGE/valid_predictions.jsonl\n", + "The evaluation finished successfully.\n", + "{\"success\": true, \"score\": 75.33280085197018, \"scoreJson\": {\"score\": 75.33280085197018, \"mean_recall\": 75.33280085197018, \"r1\": 56.86900958466453, \"r5\": 81.21006389776358, \"r10\": 87.91932907348243}}" + ] + } + ], + "source": [ + "# 根据top-10预测结果，计算验证集的Recall@1/5/10，同时给出mean recall（Recall@1/5/10的平均数）\n", + "\n", + "split=\"valid\" # 指定计算valid或test集特征\n", + "\n", + "run_command = \"export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\" + \\\n", + "f\"\"\"\n", + "python cn_clip/eval/evaluation.py \\\n", + " {DATAPATH}/datasets/{dataset_name}/{split}_texts.jsonl \\\n", + " {DATAPATH}/datasets/{dataset_name}/{split}_predictions.jsonl \\\n", + " output.json;\n", + "cat output.json\n", + "\"\"\"\n", + "print(run_command.lstrip())\n", + "!{run_command}" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "可以看到验证集上，我们取得了~75的Mean Recall值，同时也给出了详细的Recall@1/5/10. 这个分数，已经比直接使用同规模预训练ckpt进行预测（zero-shot）的71.1更高，说明我们的finetune让模型更加适配到电商检索任务的领域数据上。通过调节不同的超参，上面finetune的分数还有很大的提升空间。在我们的实验中，Chinese-CLIP各规模zero-shot以及finetune，在验证集上可以得到的分数大体如下（更多Chinese-CLIP在其他任务的实验结果，请详见Github代码库 https://github.com/OFA-Sys/Chinese-CLIP/blob/master/Results.md ）：" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "**MUGE Text-to-Image Retrieval (Official Validation Set)**:\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "\t\n", + " \n", + " \n", + "\t\n", + " \n", + " \n", + "\t\n", + " \n", + " \n", + "\t\n", + " \n", + " \n", + "\t\n", + " \n", + " \n", + "
SetupZero-shotFinetune
MetricR@1R@5R@10MRR@1R@5R@10MR
CN-CLIPRN5042.668.677.963.048.675.184.069.2
CN-CLIPViT-B/1652.176.784.471.158.483.690.077.4
CN-CLIPViT-L/1456.379.886.274.163.385.691.380.1
CN-CLIPViT-L/14@336px59.081.487.876.165.386.792.181.3
CN-CLIPViT-H/1463.084.189.278.868.988.793.183.6
\n", + "
" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "ExecutionIndicator": { + "show": false + } + }, + "source": [ + "## 准备测试集结果" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "下面我们在测试集上，再运行一次和验证集同样的特征计算和KNN召回流程，只是数据换成测试集，从而准备一份测试集预测结果，用于提交官方" + ] + }, + { + "cell_type": "code", + "execution_count": 50, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-14T13:45:44.959643Z", + "iopub.status.busy": "2023-01-14T13:45:44.959352Z", + "iopub.status.idle": "2023-01-14T13:52:18.502675Z", + "shell.execute_reply": "2023-01-14T13:52:18.501907Z", + "shell.execute_reply.started": "2023-01-14T13:45:44.959618Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\n", + "python -u cn_clip/eval/extract_features.py --extract-image-feats --extract-text-feats --image-data=\"../datapath//datasets/MUGE/lmdb/test/imgs\" --text-data=\"../datapath//datasets/MUGE/test_texts.jsonl\" --img-batch-size=32 --text-batch-size=32 --context-length=52 --resume=../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt --vision-model=ViT-B-16 --text-model=RoBERTa-wwm-ext-base-chinese\n", + "\n", + "/home/pai/lib/python3.6/site-packages/OpenSSL/crypto.py:12: CryptographyDeprecationWarning: Python 3.6 is no longer supported by the Python core team. Therefore, support for it is deprecated in cryptography and will be removed in a future release.\n", + " from cryptography import x509\n", + "Params:\n", + " context_length: 52\n", + " debug: False\n", + " extract_image_feats: True\n", + " extract_text_feats: True\n", + " image_data: ../datapath//datasets/MUGE/lmdb/test/imgs\n", + " image_feat_output_path: None\n", + " img_batch_size: 32\n", + " precision: amp\n", + " resume: ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt\n", + " text_batch_size: 32\n", + " text_data: ../datapath//datasets/MUGE/test_texts.jsonl\n", + " text_feat_output_path: None\n", + " text_model: RoBERTa-wwm-ext-base-chinese\n", + " vision_model: ViT-B-16\n", + "Loading vision model config from cn_clip/clip/model_configs/ViT-B-16.json\n", + "Loading text model config from cn_clip/clip/model_configs/RoBERTa-wwm-ext-base-chinese.json\n", + "Preparing image inference dataset.\n", + "Preparing text inference dataset.\n", + "Begin to load model checkpoint from ../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt.\n", + "=> loaded checkpoint '../datapath//experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt' (epoch 1 @ 5215 steps)\n", + "Make inference for texts...\n", + "100%|█████████████████████████████████████████| 157/157 [00:07<00:00, 21.72it/s]\n", + "5004 text features are stored in ../datapath//datasets/MUGE/test_texts.txt_feat.jsonl\n", + "Make inference for images...\n", + "100%|█████████████████████████████████████████| 950/950 [06:13<00:00, 2.54it/s]\n", + "30399 image features are stored in ../datapath//datasets/MUGE/test_imgs.img_feat.jsonl\n", + "Done!\n" + ] + } + ], + "source": [ + "# 为测试集图片池和query文本计算特征\n", + "dataset_name=\"MUGE\"\n", + "split=\"test\" # 指定计算valid或test集特征\n", + "resume=f\"{DATAPATH}/experiments/muge_finetune_vit-b-16_roberta-base_bs48_1gpu/checkpoints/epoch_latest.pt\"\n", + "\n", + "run_command = \"export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\" + \\\n", + "f\"\"\"\n", + "python -u cn_clip/eval/extract_features.py \\\n", + " --extract-image-feats \\\n", + " --extract-text-feats \\\n", + " --image-data=\"{DATAPATH}/datasets/{dataset_name}/lmdb/{split}/imgs\" \\\n", + " --text-data=\"{DATAPATH}/datasets/{dataset_name}/{split}_texts.jsonl\" \\\n", + " --img-batch-size=32 \\\n", + " --text-batch-size=32 \\\n", + " --context-length=52 \\\n", + " --resume={resume} \\\n", + " --vision-model=ViT-B-16 \\\n", + " --text-model=RoBERTa-wwm-ext-base-chinese\n", + "\"\"\"\n", + "print(run_command.lstrip())\n", + "!{run_command}" + ] + }, + { + "cell_type": "code", + "execution_count": 51, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-14T13:53:09.035737Z", + "iopub.status.busy": "2023-01-14T13:53:09.035458Z", + "iopub.status.idle": "2023-01-14T13:56:00.198365Z", + "shell.execute_reply": "2023-01-14T13:56:00.197573Z", + "shell.execute_reply.started": "2023-01-14T13:53:09.035711Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\n", + "python -u cn_clip/eval/make_topk_predictions.py --image-feats=\"../datapath//datasets/MUGE/test_imgs.img_feat.jsonl\" --text-feats=\"../datapath//datasets/MUGE/test_texts.txt_feat.jsonl\" --top-k=10 --eval-batch-size=32768 --output=\"../datapath//datasets/MUGE/test_predictions.jsonl\"\n", + "\n", + "Params:\n", + " eval_batch_size: 32768\n", + " image_feats: ../datapath//datasets/MUGE/test_imgs.img_feat.jsonl\n", + " output: ../datapath//datasets/MUGE/test_predictions.jsonl\n", + " text_feats: ../datapath//datasets/MUGE/test_texts.txt_feat.jsonl\n", + " top_k: 10\n", + "Begin to load image features...\n", + "30399it [00:07, 4257.35it/s]\n", + "Finished loading image features.\n", + "Begin to compute top-10 predictions for texts...\n", + "5004it [02:40, 31.19it/s]\n", + "Top-10 predictions are saved in ../datapath//datasets/MUGE/test_predictions.jsonl\n", + "Done!\n" + ] + } + ], + "source": [ + "# 进行KNN检索，为测试集每个query，匹配特征余弦相似度最高的top-10商品图片\n", + "\n", + "split=\"test\" # 指定计算valid或test集特征\n", + "\n", + "run_command = \"export PYTHONPATH=${PYTHONPATH}:`pwd`/cn_clip;\" + \\\n", + "f\"\"\"\n", + "python -u cn_clip/eval/make_topk_predictions.py \\\n", + " --image-feats=\"{DATAPATH}/datasets/{dataset_name}/{split}_imgs.img_feat.jsonl\" \\\n", + " --text-feats=\"{DATAPATH}/datasets/{dataset_name}/{split}_texts.txt_feat.jsonl\" \\\n", + " --top-k=10 \\\n", + " --eval-batch-size=32768 \\\n", + " --output=\"{DATAPATH}/datasets/{dataset_name}/{split}_predictions.jsonl\"\n", + "\"\"\"\n", + "print(run_command.lstrip())\n", + "!{run_command}" + ] + }, + { + "cell_type": "code", + "execution_count": 52, + "metadata": { + "execution": { + "iopub.execute_input": "2023-01-14T13:56:02.000527Z", + "iopub.status.busy": "2023-01-14T13:56:02.000133Z", + "iopub.status.idle": "2023-01-14T13:56:02.302021Z", + "shell.execute_reply": "2023-01-14T13:56:02.301051Z", + "shell.execute_reply.started": "2023-01-14T13:56:02.000481Z" + } + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "{\"text_id\": 342160, \"image_ids\": [599057, 224560, 282239, 908966, 155774, 511583, 368883, 611578, 808493, 598944]}\n", + "{\"text_id\": 342169, \"image_ids\": [686662, 191142, 344670, 1122644, 986797, 406083, 969455, 424464, 2971, 251105]}\n", + "{\"text_id\": 342177, \"image_ids\": [75011, 173661, 108900, 454071, 814416, 1114918, 331928, 474673, 98571, 1069979]}\n", + "{\"text_id\": 342198, \"image_ids\": [526032, 277415, 1015004, 443782, 612288, 862027, 944337, 780085, 1122730, 442797]}\n", + "{\"text_id\": 342202, \"image_ids\": [258830, 270450, 426942, 807698, 529722, 465381, 777309, 937790, 650457, 23176]}\n" + ] + } + ], + "source": [ + "!head -n 5 ../datapath/datasets/MUGE/test_predictions.jsonl" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "该结果下载并提交到天池后，预期将取得56.5/80.6/87.5左右的Recall@1/5/10，即74.9左右的Mean Recall，可能由于finetune的随机性有一些小幅度的浮动" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "以上就是我们使用Chinese-CLIP，完成电商图文检索挑战赛的一个完整的baseline流程，欢迎大家在此基础上进一步提升效果。也希望大家多多支持Chinese-CLIP，如果觉得有帮助请为我们的 **[Github repo](https://github.com/OFA-Sys/Chinese-CLIP)** 点点star⭐️，多提宝贵的意见！" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.6.15" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/MIT-LICENSE.txt b/MIT-LICENSE.txt index 4b3fa65..fec2835 100644 --- a/MIT-LICENSE.txt +++ b/MIT-LICENSE.txt @@ -1,4 +1,4 @@ -Copyright (c) 2012-2022 OFA-Sys Team +Copyright (c) 2022-2023 OFA-Sys Team Copyright (c) 2012-2022 Gabriel Ilharco, Mitchell Wortsman, Nicholas Carlini, Rohan Taori, Achal Dave, Vaishaal Shankar, John Miller, Hongseok Namkoong, Hannaneh Hajishirzi, Ali Farhadi, Ludwig Schmidt diff --git a/README.md b/README.md index 48014b1..d9f881f 100644 --- a/README.md +++ b/README.md @@ -8,7 +8,7 @@

- ModelScope  ｜  Demo  ｜  Paper  ｜  Blog + ModelScope  ｜  Demo  ｜  Paper  ｜  Blog



@@ -39,25 +39,28 @@ Chinese-CLIP目前开源5个不同规模，其模型信息和下载方式见下 - + - + - + - + - + - +
模型规模下载链接参数量视觉侧骨架视觉侧参数量文本侧骨架文本侧参数量分辨率模型名称模型规模下载链接参数量视觉侧骨架视觉侧参数量文本侧骨架文本侧参数量分辨率
CN-CLIPRN50Download77MResNet5038MRBT339M224chinese-clip-rn50CN-CLIPRN50🤗 🤖77MResNet5038MRBT339M224
CN-CLIPViT-B/16Download188MViT-B/1686MRoBERTa-wwm-Base102M224chinese-clip-vit-base-patch16CN-CLIPViT-B/16🤗 🤖188MViT-B/1686MRoBERTa-wwm-Base102M224
CN-CLIPViT-L/14Download406MViT-L/14304MRoBERTa-wwm-Base102M224chinese-clip-vit-large-patch14CN-CLIPViT-L/14🤗 🤖406MViT-L/14304MRoBERTa-wwm-Base102M224
CN-CLIPViT-L/14@336pxDownload407MViT-L/14304MRoBERTa-wwm-Base102M336chinese-clip-vit-large-patch14-336pxCN-CLIPViT-L/14@336px🤗 🤖407MViT-L/14304MRoBERTa-wwm-Base102M336
CN-CLIPViT-H/14Download958MViT-H/14632MRoBERTa-wwm-Large326M224chinese-clip-vit-huge-patch14CN-CLIPViT-H/14🤗 🤖958MViT-H/14632MRoBERTa-wwm-Large326M224
-

+ +- 🤗 Hugging Face Hub +- 🤖 魔搭社区 + ## 实验结果 针对图文检索任务，我们在[MUGE Retrieval](https://tianchi.aliyun.com/muge)、[Flickr30K-CN](https://github.com/li-xirong/cross-lingual-cap)和[COCO-CN](https://github.com/li-xirong/coco-cn)上进行了zero-shot和finetune的实验。针对图像零样本分类，我们在[ELEVATER](https://eval.ai/web/challenges/challenge-page/1832)的10个数据集上进行了实验。实验结果如下表所示。篇幅所限，我们这里给出baseline模型和Chinese-CLIP的最优规模模型结果，关于Chinese-CLIP各规模的详细结果指标，请详见[Results.md](Results.md)。 @@ -194,7 +197,8 @@ print("Available models:", available_models()) # Available models: ['ViT-B-16', 'ViT-L-14', 'ViT-L-14-336', 'ViT-H-14', 'RN50'] device = "cuda" if torch.cuda.is_available() else "cpu" -model, preprocess = load_from_name("ViT-B-16", device=device, download_root='./') +# 如本地模型不存在，自动从ModelScope下载模型，需要提前安装`modelscope`包 +model, preprocess = load_from_name("ViT-B-16", device=device, download_root='./', use_modelscope=True) model.eval() image = preprocess(Image.open("examples/pokemon.jpeg")).unsqueeze(0).to(device) text = clip.tokenize(["杰尼龟", "妙蛙种子", "小火龙", "皮卡丘"]).to(device) @@ -251,7 +255,7 @@ ${DATAPATH} #### 预训练CKPT -请参考前文[模型规模 & 下载链接](#model_card)部分，下载对应模型ckpt。推荐将下载的ckpt文件存放于`${DATAPATH}/pretrained_weights/`目录下。 +请参考前文[模型规模 & 下载链接](#model_card)部分，下载对应模型ckpt（`.pt`后缀文件即可）。推荐将下载的ckpt文件存放于`${DATAPATH}/pretrained_weights/`目录下。 #### 数据集格式预处理 @@ -314,7 +318,7 @@ ${DATAPATH} └── test ``` -为了降低上手难度，我们也提供了按上述步骤预处理好的MUGE数据（[下载链接](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/MUGE.zip)）和Flickr30K-CN数据（[下载链接](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/Flickr30k-CN.zip)）压缩包，直接下载解压并放置于`${DATAPATH}/datasets/`目录下即可。如果需要[COCO-CN](https://github.com/li-xirong/coco-cn)数据，请向原作者进行申请许可完成后，邮件联系我们吧。 +为了降低上手难度，我们也提供了按上述步骤预处理好的MUGE数据（[🤗下载链接](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/blob/main/MUGE.zip)）和Flickr30K-CN数据（[🤗下载链接](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/resolve/main/Flickr30k-CN.zip)）压缩包，直接下载解压并放置于`${DATAPATH}/datasets/`目录下即可。如果需要[COCO-CN](https://github.com/li-xirong/coco-cn)数据，请向原作者进行申请许可完成后，邮件联系我们吧。 ### 模型finetune @@ -478,7 +482,7 @@ cat output.json {"success": true, "score": 85.67, "scoreJson": {"score": 85.67, "mean_recall": 85.67, "r1": 71.2, "r5": 90.5, "r10": 95.3}} ``` -关于整套跨模态检索的训练和测试流程，我们以MUGE检索数据集（[多模态电商图文挑战赛](https://tianchi.aliyun.com/competition/entrance/532031/introduction)）为例，也提供了一个包含上述全部流程并可运行的Jupyter Notebook（[下载链接](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/others/Chinese-CLIP-on-MUGE-Retrieval.ipynb)），欢迎大家上手实践。 +关于整套跨模态检索的训练和测试流程，我们以MUGE检索数据集（[多模态电商图文挑战赛](https://tianchi.aliyun.com/competition/entrance/532031/introduction)）为例，也提供了一个包含上述全部流程并可运行的Jupyter Notebook（[下载链接](Chinese-CLIP-on-MUGE-Retrieval.ipynb)），欢迎大家上手实践。
@@ -517,7 +521,7 @@ ${DATAPATH} 锚 ... ``` -每行的标签对应的label id为`行号-1`，如第1行的标签的id为0，第二行的标签的id为1。如果标签总数大于10，则统一向左补零到3位数，比如标签个数为100，标签id则为`000-099`。用户需为每个label id生成对应的文件夹，并将标注该label的样本放入其中。我们以ELEVATER中的**CIFAR-100数据集**为样例，请点击[链接](http://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/cifar-100.zip)下载处理好的数据。如果想尝试在其他ELEVATER包含的数据集上测试Chinese-CLIP，请参见我们的[数据文档](https://github.com/OFA-Sys/Chinese-CLIP/blob/master/zeroshot_dataset.md)。 +每行的标签对应的label id为`行号-1`，如第1行的标签的id为0，第二行的标签的id为1。如果标签总数大于10，则统一向左补零到3位数，比如标签个数为100，标签id则为`000-099`。用户需为每个label id生成对应的文件夹，并将标注该label的样本放入其中。我们以ELEVATER中的**CIFAR-100数据集**为样例，请访问数据集[🤗OFA-Sys/chinese-clip-eval](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval)或点击[🤗链接](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/resolve/main/cifar-100.zip)下载处理好的数据。如果想尝试在其他ELEVATER包含的数据集上测试Chinese-CLIP，请参见我们的[数据文档](https://github.com/OFA-Sys/Chinese-CLIP/blob/master/zeroshot_dataset.md)。
### 预测和评估 diff --git a/README_En.md b/README_En.md index 583253c..dab6c20 100644 --- a/README_En.md +++ b/README_En.md @@ -8,7 +8,7 @@

- ModelScope  ｜  Demo  ｜  Paper   ｜  Blog + ModelScope  ｜  Demo  ｜  Paper  ｜  Blog



@@ -39,25 +39,27 @@ Currently, we release 5 different sizes of Chinese-CLIP models. Detailed informa - + - + - + - + - + - +
ModelCkpt#Params (All)Backbone (I)#Params (I)Backbone (T)#Params (T)ResolutionModel IDModelCkpt#Params (All)Backbone (I)#Params (I)Backbone (T)#Params (T)Resolution
CN-CLIPRN50Download77MResNet5038MRBT339M224chinese-clip-rn50CN-CLIPRN50🤗 🤖77MResNet5038MRBT339M224
CN-CLIPViT-B/16Download188MViT-B/1686MRoBERTa-wwm-Base102M224chinese-clip-vit-base-patch16CN-CLIPViT-B/16🤗 🤖188MViT-B/1686MRoBERTa-wwm-Base102M224
CN-CLIPViT-L/14Download406MViT-L/14304MRoBERTa-wwm-Base102M224chinese-clip-vit-large-patch14CN-CLIPViT-L/14🤗 🤖406MViT-L/14304MRoBERTa-wwm-Base102M224
CN-CLIPViT-L/14@336pxDownload407MViT-L/14304MRoBERTa-wwm-Base102M336chinese-clip-vit-large-patch14-336pxCN-CLIPViT-L/14@336px🤗 🤖407MViT-L/14304MRoBERTa-wwm-Base102M336
CN-CLIPViT-H/14Download958MViT-H/14632MRoBERTa-wwm-Large326M224chinese-clip-vit-huge-patch14CN-CLIPViT-H/14🤗 🤖958MViT-H/14632MRoBERTa-wwm-Large326M224
-

+ +- 🤗 Hugging Face Hub +- 🤖 ModelScope ## Results We conducted zero-shot inference and finetuning experiments on [MUGE Retrieval](https://tianchi.aliyun.com/muge), [Flickr30K-CN](https://github.com/li-xirong/cross-lingual-cap) and [COCO-CN](https://github.com/li-xirong/coco-cn) for the evaluation of cross-modal retrieval, and conducted experiments on 10 image classification datasets of the [ELEVATER](https://eval.ai/web/challenges/challenge-page/1832) benchmark for the evaluation of zero-shot image classification. Results are shown below. Due to space limitation, here we only list the performance of the best performing Chinese-CLIP and baseline models. For detailed performance of each Chinese-CLIP model size, please refer to [Results.md](Results.md). @@ -192,6 +194,8 @@ print("Available models:", available_models()) # Available models: ['ViT-B-16', 'ViT-L-14', 'ViT-L-14-336', 'ViT-H-14', 'RN50'] device = "cuda" if torch.cuda.is_available() else "cpu" +# If the model checkpoint does not exist in local drives, it will download the checkpoint from Hugging Face Hub automatically. +# Requires `huggingface_hub` to be installed. model, preprocess = load_from_name("ViT-B-16", device=device, download_root='./') model.eval() image = preprocess(Image.open("examples/pokemon.jpeg")).unsqueeze(0).to(device) @@ -315,7 +319,7 @@ ${DATAPATH} └── test ``` -For easier use, we have provided preprocessed MUGE ([download link](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/MUGE.zip)) and Flickr30K-CN ([download link](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/Flickr30k-CN.zip)) datasets in zip format. To use them, just download and unzip it under `${DATAPATH}/datasets/`. If you need [COCO-CN](https://github.com/li-xirong/coco-cn) dataset, please contact us by email when you have finished applying for permission from the original author. +For easier use, we have provided preprocessed MUGE ([🤗download link](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/resolve/main/MUGE.zip)) and Flickr30K-CN ([🤗download link](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/resolve/main/Flickr30k-CN.zip)) datasets in zip format. To use them, just download and unzip it under `${DATAPATH}/datasets/`. If you need [COCO-CN](https://github.com/li-xirong/coco-cn) dataset, please contact us by email when you have finished applying for permission from the original author. ### Finetuning @@ -481,7 +485,7 @@ The printed results are shown below: {"success": true, "score": 85.67, "scoreJson": {"score": 85.67, "mean_recall": 85.67, "r1": 71.2, "r5": 90.5, "r10": 95.3}} ``` -For better understanding of cross-modal retrieval by Chinese-CLIP, we also provide a runnable jupyter notebook ([download link](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/others/Chinese-CLIP-on-MUGE-Retrieval.ipynb)), which works with the MUGE retrieval dataset (corresponding leaderboard is hosted on [Tianchi](https://tianchi.aliyun.com/competition/entrance/532031/introduction?lang=en-us)) and includes the finetuning and inference process mentioned above. Welcome to try! +For better understanding of cross-modal retrieval by Chinese-CLIP, we also provide a runnable jupyter notebook ([download link](Chinese-CLIP-on-MUGE-Retrieval.ipynb)), which works with the MUGE retrieval dataset (corresponding leaderboard is hosted on [Tianchi](https://tianchi.aliyun.com/competition/entrance/532031/introduction?lang=en-us)) and includes the finetuning and inference process mentioned above. Welcome to try!
@@ -520,7 +524,7 @@ airplane anchor ... ``` -The label id is `[line number]-1`. For example, the label id for the first line is 0, and the one for the second line is 1. If the number of labels is larger than 10, all labels are filled with 0 by the left to 3-digit numbers. For example, if the number of labels is 100, the ids are `000-099`. Users should create a directory for each label, and put the corresponding samples into the directories. We provide the processed dataset CIFAR-100 as an example, and please click [this link](http://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/cifar-100.zip) to download the prepared dataset. To evaluate other datasets of ELEVATER, please refer to [Notes for datasets](zeroshot_dataset_en.md) for download. +The label id is `[line number]-1`. For example, the label id for the first line is 0, and the one for the second line is 1. If the number of labels is larger than 10, all labels are filled with 0 by the left to 3-digit numbers. For example, if the number of labels is 100, the ids are `000-099`. Users should create a directory for each label, and put the corresponding samples into the directories. We provide the processed dataset CIFAR-100 as an example, and please visit dataset [🤗 OFA-Sys/chinese-clip-eval](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval) or click [🤗 link](https://huggingface.co/datasets/OFA-Sys/chinese-clip-eval/resolve/main/cifar-100.zip) to download the prepared dataset. To evaluate other datasets of ELEVATER, please refer to [Notes for datasets](zeroshot_dataset_en.md) for download.

### Prediction and Evaluation diff --git a/cn_clip/clip/utils.py b/cn_clip/clip/utils.py index 0b7367a..6defa7a 100644 --- a/cn_clip/clip/utils.py +++ b/cn_clip/clip/utils.py @@ -4,24 +4,26 @@ import os from pathlib import Path from typing import Union, List -import urllib import torch from torchvision.transforms import Compose, ToTensor, Normalize, Resize, InterpolationMode -from tqdm import tqdm from cn_clip.clip import _tokenizer from cn_clip.clip.model import convert_weights, CLIP, restore_model __all__ = ["load", "tokenize", "available_models", "image_transform", "load_from_name"] +_MODELSCOPE_ORG = "AI-ModelScope" +_HUGGINGFACE_ORG = "OFA-Sys" + _MODELS = { - "ViT-B-16": "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/checkpoints/clip_cn_vit-b-16.pt", - "ViT-L-14": "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/checkpoints/clip_cn_vit-l-14.pt", - "ViT-L-14-336": "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/checkpoints/clip_cn_vit-l-14-336.pt", - "ViT-H-14": "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/checkpoints/clip_cn_vit-h-14.pt", - "RN50": "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/checkpoints/clip_cn_rn50.pt", + "ViT-B-16": ("chinese-clip-vit-base-patch16", "clip_cn_vit-b-16.pt"), + "ViT-L-14": ("chinese-clip-vit-large-patch14", "clip_cn_vit-l-14.pt"), + "ViT-L-14-336": ("chinese-clip-vit-large-patch14-336px", "clip_cn_vit-l-14-336.pt"), + "ViT-H-14": ("chinese-clip-vit-huge-patch14", "clip_cn_vit-h-14.pt"), + "RN50": ("chinese-clip-rn50", "clip_cn_rn50.pt"), } + _MODEL_INFO = { "ViT-B-16": { "struct": "ViT-B-16@RoBERTa-wwm-ext-base-chinese", @@ -46,9 +48,11 @@ } -def _download(url: str, root: str): +def _download(modelname: str, root: str, use_modelscope: bool = False): os.makedirs(root, exist_ok=True) - filename = os.path.basename(url) + + # this a private function and the only caller has checked the model exsits + reponame, filename = _MODELS[modelname] download_target = os.path.join(root, filename) @@ -58,18 +62,36 @@ def _download(url: str, root: str): if os.path.isfile(download_target): return download_target - with urllib.request.urlopen(url) as source, open(download_target, "wb") as output: - with tqdm(total=int(source.info().get("Content-Length")), ncols=80, unit='iB', unit_scale=True, - unit_divisor=1024) as loop: - while True: - buffer = source.read(8192) - if not buffer: - break + if use_modelscope: + try: + from modelscope.hub.file_download import model_file_download + except ImportError as _: + raise RuntimeError( + "Ckpt download requires `modelscope`. " + "Please install `modelscope` or download the ckpt manually and " + "provide the local path so that we can continue." + ) + + local_path = model_file_download( + model_id=_MODELSCOPE_ORG + "/" + reponame, + file_path=filename, + local_dir=root, + ) + else: + try: + from huggingface_hub import hf_hub_download + except ImportError as _: + raise RuntimeError( + "Ckpt download requires `huggingface_hub`. " + "Please install `huggingface_hub` or download the ckpt manually and " + "provide the local path so that we can continue." + ) - output.write(buffer) - loop.update(len(buffer)) + local_path = hf_hub_download( + _HUGGINGFACE_ORG + "/" + reponame, filename=filename, local_dir=root + ) - return download_target + return local_path def _convert_image_to_rgb(image): @@ -82,9 +104,9 @@ def available_models() -> List[str]: def load_from_name(name: str, device: Union[str, torch.device] = "cuda" if torch.cuda.is_available() else "cpu", - download_root: str = None, vision_model_name: str = None, text_model_name: str = None, input_resolution: int = None): + download_root: str = None, vision_model_name: str = None, text_model_name: str = None, input_resolution: int = None, use_modelscope: bool = False): if name in _MODELS: - model_path = _download(_MODELS[name], download_root or os.path.expanduser("~/.cache/clip")) + model_path = _download(name, download_root or os.path.expanduser("~/.cache/clip"), use_modelscope) model_name, model_input_resolution = _MODEL_INFO[name]['struct'], _MODEL_INFO[name]['input_resolution'] elif os.path.isfile(name): assert vision_model_name and text_model_name and input_resolution, "Please specify specific 'vision_model_name', 'text_model_name', and 'input_resolution'" diff --git a/datasets/ImageNet-1K/label.txt b/datasets/ImageNet-1K/label.txt new file mode 100644 index 0000000..bfcfa7b --- /dev/null +++ b/datasets/ImageNet-1K/label.txt @@ -0,0 +1,1000 @@ +tench +goldfish +great white shark +tiger shark +hammerhead shark +electric ray +stingray +rooster +hen +ostrich +brambling +goldfinch +house finch +junco +indigo bunting +American robin +bulbul +jay +magpie +chickadee +American dipper +kite (bird of prey) +bald eagle +vulture +great grey owl +fire salamander +smooth newt +newt +spotted salamander +axolotl +American bullfrog +tree frog +tailed frog +loggerhead sea turtle +leatherback sea turtle +mud turtle +terrapin +box turtle +banded gecko +green iguana +Carolina anole +desert grassland whiptail lizard +agama +frilled-necked lizard +alligator lizard +Gila monster +European green lizard +chameleon +Komodo dragon +Nile crocodile +American alligator +triceratops +worm snake +ring-necked snake +eastern hog-nosed snake +smooth green snake +kingsnake +garter snake +water snake +vine snake +night snake +boa constrictor +African rock python +Indian cobra +green mamba +sea snake +Saharan horned viper +eastern diamondback rattlesnake +sidewinder rattlesnake +trilobite +harvestman +scorpion +yellow garden spider +barn spider +European garden spider +southern black widow +tarantula +wolf spider +tick +centipede +black grouse +ptarmigan +ruffed grouse +prairie grouse +peafowl +quail +partridge +african grey parrot +macaw +sulphur-crested cockatoo +lorikeet +coucal +bee eater +hornbill +hummingbird +jacamar +toucan +duck +red-breasted merganser +goose +black swan +tusker +echidna +platypus +wallaby +koala +wombat +jellyfish +sea anemone +brain coral +flatworm +nematode +conch +snail +slug +sea slug +chiton +chambered nautilus +Dungeness crab +rock crab +fiddler crab +red king crab +American lobster +spiny lobster +crayfish +hermit crab +isopod +white stork +black stork +spoonbill +flamingo +little blue heron +great egret +bittern bird +crane bird +limpkin +common gallinule +American coot +bustard +ruddy turnstone +dunlin +common redshank +dowitcher +oystercatcher +pelican +king penguin +albatross +grey whale +killer whale +dugong +sea lion +Chihuahua +Japanese Chin +Maltese +Pekingese +Shih Tzu +King Charles Spaniel +Papillon +toy terrier +Rhodesian Ridgeback +Afghan Hound +Basset Hound +Beagle +Bloodhound +Bluetick Coonhound +Black and Tan Coonhound +Treeing Walker Coonhound +English foxhound +Redbone Coonhound +borzoi +Irish Wolfhound +Italian Greyhound +Whippet +Ibizan Hound +Norwegian Elkhound +Otterhound +Saluki +Scottish Deerhound +Weimaraner +Staffordshire Bull Terrier +American Staffordshire Terrier +Bedlington Terrier +Border Terrier +Kerry Blue Terrier +Irish Terrier +Norfolk Terrier +Norwich Terrier +Yorkshire Terrier +Wire Fox Terrier +Lakeland Terrier +Sealyham Terrier +Airedale Terrier +Cairn Terrier +Australian Terrier +Dandie Dinmont Terrier +Boston Terrier +Miniature Schnauzer +Giant Schnauzer +Standard Schnauzer +Scottish Terrier +Tibetan Terrier +Australian Silky Terrier +Soft-coated Wheaten Terrier +West Highland White Terrier +Lhasa Apso +Flat-Coated Retriever +Curly-coated Retriever +Golden Retriever +Labrador Retriever +Chesapeake Bay Retriever +German Shorthaired Pointer +Vizsla +English Setter +Irish Setter +Gordon Setter +Brittany dog +Clumber Spaniel +English Springer Spaniel +Welsh Springer Spaniel +Cocker Spaniel +Sussex Spaniel +Irish Water Spaniel +Kuvasz +Schipperke +Groenendael dog +Malinois +Briard +Australian Kelpie +Komondor +Old English Sheepdog +Shetland Sheepdog +collie +Border Collie +Bouvier des Flandres dog +Rottweiler +German Shepherd Dog +Dobermann +Miniature Pinscher +Greater Swiss Mountain Dog +Bernese Mountain Dog +Appenzeller Sennenhund +Entlebucher Sennenhund +Boxer +Bullmastiff +Tibetan Mastiff +French Bulldog +Great Dane +St. Bernard +husky +Alaskan Malamute +Siberian Husky +Dalmatian +Affenpinscher +Basenji +pug +Leonberger +Newfoundland dog +Great Pyrenees dog +Samoyed +Pomeranian +Chow Chow +Keeshond +brussels griffon +Pembroke Welsh Corgi +Cardigan Welsh Corgi +Toy Poodle +Miniature Poodle +Standard Poodle +Mexican hairless dog (xoloitzcuintli) +grey wolf +Alaskan tundra wolf +red wolf or maned wolf +coyote +dingo +dhole +African wild dog +hyena +red fox +kit fox +Arctic fox +grey fox +tabby cat +tiger cat +Persian cat +Siamese cat +Egyptian Mau +cougar +lynx +leopard +snow leopard +jaguar +lion +tiger +cheetah +brown bear +American black bear +polar bear +sloth bear +mongoose +meerkat +tiger beetle +ladybug +ground beetle +longhorn beetle +leaf beetle +dung beetle +rhinoceros beetle +weevil +fly +bee +ant +grasshopper +cricket insect +stick insect +cockroach +praying mantis +cicada +leafhopper +lacewing +dragonfly +damselfly +red admiral butterfly +ringlet butterfly +monarch butterfly +small white butterfly +sulphur butterfly +gossamer-winged butterfly +starfish +sea urchin +sea cucumber +cottontail rabbit +hare +Angora rabbit +hamster +porcupine +fox squirrel +marmot +beaver +guinea pig +common sorrel horse +zebra +pig +wild boar +warthog +hippopotamus +ox +water buffalo +bison +ram (adult male sheep) +bighorn sheep +Alpine ibex +hartebeest +impala (antelope) +gazelle +arabian camel +llama +weasel +mink +European polecat +black-footed ferret +otter +skunk +badger +armadillo +three-toed sloth +orangutan +gorilla +chimpanzee +gibbon +siamang +guenon +patas monkey +baboon +macaque +langur +black-and-white colobus +proboscis monkey +marmoset +white-headed capuchin +howler monkey +titi monkey +Geoffroy's spider monkey +common squirrel monkey +ring-tailed lemur +indri +Asian elephant +African bush elephant +red panda +giant panda +snoek fish +eel +silver salmon +rock beauty fish +clownfish +sturgeon +gar fish +lionfish +pufferfish +abacus +abaya +academic gown +accordion +acoustic guitar +aircraft carrier +airliner +airship +altar +ambulance +amphibious vehicle +analog clock +apiary +apron +trash can +assault rifle +backpack +bakery +balance beam +balloon +ballpoint pen +Band-Aid +banjo +baluster / handrail +barbell +barber chair +barbershop +barn +barometer +barrel +wheelbarrow +baseball +basketball +bassinet +bassoon +swimming cap +bath towel +bathtub +station wagon +lighthouse +beaker +military hat (bearskin or shako) +beer bottle +beer glass +bell tower +baby bib +tandem bicycle +bikini +ring binder +binoculars +birdhouse +boathouse +bobsleigh +bolo tie +poke bonnet +bookcase +bookstore +bottle cap +hunting bow +bow tie +brass memorial plaque +bra +breakwater +breastplate +broom +bucket +buckle +bulletproof vest +high-speed train +butcher shop +taxicab +cauldron +candle +cannon +canoe +can opener +cardigan +car mirror +carousel +tool kit +cardboard box / carton +car wheel +automated teller machine +cassette +cassette player +castle +catamaran +CD player +cello +mobile phone +chain +chain-link fence +chain mail +chainsaw +storage chest +chiffonier +bell or wind chime +china cabinet +Christmas stocking +church +movie theater +cleaver +cliff dwelling +cloak +clogs +cocktail shaker +coffee mug +coffeemaker +spiral or coil +combination lock +computer keyboard +candy store +container ship +convertible +corkscrew +cornet +cowboy boot +cowboy hat +cradle +construction crane +crash helmet +crate +infant bed +Crock Pot +croquet ball +crutch +cuirass +dam +desk +desktop computer +rotary dial telephone +diaper +digital clock +digital watch +dining table +dishcloth +dishwasher +disc brake +dock +dog sled +dome +doormat +drilling rig +drum +drumstick +dumbbell +Dutch oven +electric fan +electric guitar +electric locomotive +entertainment center +envelope +espresso machine +face powder +feather boa +filing cabinet +fireboat +fire truck +fire screen +flagpole +flute +folding chair +football helmet +forklift +fountain +fountain pen +four-poster bed +freight car +French horn +frying pan +fur coat +garbage truck +gas mask or respirator +gas pump +goblet +go-kart +golf ball +golf cart +gondola +gong +gown +grand piano +greenhouse +radiator grille +grocery store +guillotine +hair clip +hair spray +half-track +hammer +hamper +hair dryer +hand-held computer +handkerchief +hard disk drive +harmonica +harp +combine harvester +hatchet +holster +home theater +honeycomb +hook +hoop skirt +gymnastic horizontal bar +horse-drawn vehicle +hourglass +iPod +clothes iron +carved pumpkin +jeans +jeep +T-shirt +jigsaw puzzle +rickshaw +joystick +kimono +knee pad +knot +lab coat +ladle +lampshade +laptop computer +lawn mower +lens cap +letter opener +library +lifeboat +lighter +limousine +ocean liner +lipstick +slip-on shoe +lotion +music speaker +loupe magnifying glass +sawmill +magnetic compass +messenger bag +mailbox +tights +one-piece bathing suit +manhole cover +maraca +marimba +mask +matchstick +maypole +maze +measuring cup +medicine cabinet +megalith +microphone +microwave oven +military uniform +milk can +minibus +miniskirt +minivan +missile +mitten +mixing bowl +mobile home +ford model t +modem +monastery +monitor +moped +mortar and pestle +graduation cap +mosque +mosquito net +vespa +mountain bike +tent +computer mouse +mousetrap +moving van +muzzle +metal nail +neck brace +necklace +baby pacifier +notebook computer +obelisk +oboe +ocarina +odometer +oil filter +pipe organ +oscilloscope +overskirt +bullock cart +oxygen mask +product packet / packaging +paddle +paddle wheel +padlock +paintbrush +pajamas +palace +pan flute +paper towel +parachute +parallel bars +park bench +parking meter +railroad car +patio +payphone +pedestal +pencil case +pencil sharpener +perfume +Petri dish +photocopier +plectrum +Pickelhaube +picket fence +pickup truck +pier +piggy bank +pill bottle +pillow +ping-pong ball +pinwheel +pirate ship +drink pitcher +block plane +planetarium +plastic bag +plate rack +farm plow +plunger +Polaroid camera +pole +police van +poncho +pool table +soda bottle +plant pot +potter's wheel +power drill +prayer rug +printer +prison +missile +projector +hockey puck +punching bag +purse +quill +quilt +race car +racket +radiator +radio +radio telescope +rain barrel +recreational vehicle +fishing casting reel +reflex camera +refrigerator +remote control +restaurant +revolver +rifle +rocking chair +rotisserie +eraser +rugby ball +ruler measuring stick +sneaker +safe +safety pin +salt shaker +sandal +sarong +saxophone +scabbard +weighing scale +school bus +schooner +scoreboard +CRT monitor +screw +screwdriver +seat belt +sewing machine +shield +shoe store +shoji screen / room divider +shopping basket +shopping cart +shovel +shower cap +shower curtain +ski +balaclava ski mask +sleeping bag +slide rule +sliding door +slot machine +snorkel +snowmobile +snowplow +soap dispenser +soccer ball +sock +solar thermal collector +sombrero +soup bowl +keyboard space bar +space heater +space shuttle +spatula +motorboat +spider web +spindle +sports car +spotlight +stage +steam locomotive +through arch bridge +steel drum +stethoscope +scarf +stone wall +stopwatch +stove +strainer +tram +stretcher +couch +stupa +submarine +suit +sundial +sunglasses +sunglasses +sunscreen +suspension bridge +mop +sweatshirt +swim trunks / shorts +swing +electrical switch +syringe +table lamp +tank +tape player +teapot +teddy bear +television +tennis ball +thatched roof +front curtain +thimble +threshing machine +throne +tile roof +toaster +tobacco shop +toilet seat +torch +totem pole +tow truck +toy store +tractor +semi-trailer truck +tray +trench coat +tricycle +trimaran +tripod +triumphal arch +trolleybus +trombone +hot tub +turnstile +typewriter keyboard +umbrella +unicycle +upright piano +vacuum cleaner +vase +vaulted or arched ceiling +velvet fabric +vending machine +vestment +viaduct +violin +volleyball +waffle iron +wall clock +wallet +wardrobe +military aircraft +sink +washing machine +water bottle +water jug +water tower +whiskey jug +whistle +hair wig +window screen +window shade +Windsor tie +wine bottle +airplane wing +wok +wooden spoon +wool +split-rail fence +shipwreck +sailboat +yurt +website +comic book +crossword +traffic or street sign +traffic light +dust jacket +menu +plate +guacamole +consomme +hot pot +trifle +ice cream +popsicle +baguette +bagel +pretzel +cheeseburger +hot dog +mashed potatoes +cabbage +broccoli +cauliflower +zucchini +spaghetti squash +acorn squash +butternut squash +cucumber +artichoke +bell pepper +cardoon +mushroom +Granny Smith apple +strawberry +orange +lemon +fig +pineapple +banana +jackfruit +cherimoya (custard apple) +pomegranate +hay +carbonara +chocolate syrup +dough +meatloaf +pizza +pot pie +burrito +red wine +espresso +tea cup +eggnog +mountain +bubble +cliff +coral reef +geyser +lakeshore +promontory +sandbar +beach +valley +volcano +baseball player +bridegroom +scuba diver +rapeseed +daisy +yellow lady's slipper +corn +acorn +rose hip +horse chestnut seed +coral fungus +agaric +gyromitra +stinkhorn mushroom +earth star fungus +hen of the woods mushroom +bolete +corn cob +toilet paper diff --git a/datasets/ImageNet-1K/label_cn.txt b/datasets/ImageNet-1K/label_cn.txt new file mode 100644 index 0000000..bdf1ed8 --- /dev/null +++ b/datasets/ImageNet-1K/label_cn.txt @@ -0,0 +1,1000 @@ +丁鲷 +金鱼 +大白鲨 +虎鲨 +锤头鲨 +电鳐 +黄貂鱼 +公鸡 +母鸡 +鸵鸟 +花鸡 +金翅雀 +红麻雀 +暗眼灯草鹀 +靛蓝彩鹀 +旅鸫 +红耳鹎 +冠蓝鸦 +喜鹊 +山雀 +美洲河乌 +鸢 +秃头鹰 +秃鹫 +大灰猫头鹰 +欧洲火蝾螈 +欧洲蝾螈 +蝾螈 +斑点钝口螈 +六角恐龙 +美国牛蛙 +树蛙 +尾蟾 +红海龟 +棱皮龟 +东方动胸龟 +钻纹龟 +箱龟 +豹紋壁虎 +绿鬣蜥 +安禄蜥 +沙漠鞭尾蜥蜴 +彩虹鬣蜥 +伞蜥蜴 +鳄蜥 +美國毒蜥 +欧洲绿蜥蜴 +变色龙 +科莫多巨蜥 +尼罗鳄 +美国短吻鳄 +三角龙 +钩盲蛇 +环颈蛇 +猪鼻蛇 +滑鳞绿树蛇 +王蛇 +袜带蛇 +水蛇 +绿瘦蛇 +夜蛇 +紅尾蚺 +非洲岩蟒 +印度眼镜蛇 +绿曼巴蛇 +海蛇 +非洲角蝰蛇 +东部菱背响尾蛇 +角响尾蛇 +三叶虫 +盲蜘蛛 +蝎子 +黑金花园蜘蛛 +谷仓蜘蛛 +十字园蛛 +黑寡妇蜘蛛 +塔兰托毒蛛 +狼蛛 +蜱 +蜈蚣 +黑琴鸡 +雷鸟 +披肩榛鸡 +草原松鸡 +孔雀 +鹌鹑 +鹧鸪 +非洲灰鹦鹉 +金刚鹦鹉 +葵花凤头鹦鹉 +虹彩吸蜜鹦鹉 +大毛鸡 +蜂虎 +犀鸟 +蜂鸟 +鹟䴕 +巨嘴鸟 +鸭子 +红胸秋沙鸭 +鹅 +黑天鹅 +巨齿动物 +澳洲针鼹 +鸭嘴兽 +沙袋鼠 +考拉 +袋熊 +水母 +海葵 +脑珊瑚 +扁虫 +线虫 +海螺 +蜗牛 +鼻涕虫 +海蛞蝓 +石鳖 +鹦鹉螺 +珍宝蟹 +红黄道蟹 +招潮蟹 +阿拉斯加帝王蟹 +美国龙虾 +断沟龙虾 +小龙虾 +寄居蟹 +大王具足虫 +白鹳 +黑鹳 +玫瑰琵鹭 +火烈鸟 +小蓝鹭 +大白鹭 +美洲麻鳽 +鹤 +秧鹤 +美洲黑水鸡 +美洲骨顶 +鸨 +红翻石鹬 +红背鹬 +红脚鹬 +长嘴半蹼鹬 +蛎鹬 +鹈鹕 +国王企鹅 +信天翁 +灰鲸 +杀人鲸 +儒艮 +海狮 +吉娃娃 +日本狆犬 +马尔济斯犬 +京巴犬 +西施犬 +查理士王小猎犬 +蝴蝶犬 +英国玩具梗 +罗得西亚长背猎狗 +阿富汗猎犬 +巴吉度猎犬 +比格犬 +寻血猎犬 +布鲁泰克猎浣熊犬 +黑褐猎浣熊犬 +树丛浣熊猎犬 +英国猎狐犬 +红浣熊猎犬 +苏俄牧羊犬 +爱尔兰猎狼犬 +意大利灰狗 +惠比特犬 +伊比莎猎犬 +挪威黑猎鹿犬 +奥达猎犬 +萨路基猎犬 +苏格兰猎鹿犬 +威玛猎犬 +斯塔福德郡斗牛犬 +美国斯塔福德郡梗 +贝德灵顿梗 +边境梗 +凯丽蓝梗 +爱尔兰梗 +诺福克梗 +诺维奇梗 +约克夏梗犬 +刚毛猎狐梗 +湖畔梗 +锡利哈姆梗 +艾尔谷犬 +凯恩梗 +澳大利亚梗 +丹迪丁蒙梗 +波士顿梗 +迷你雪纳瑞犬 +巨型雪纳瑞犬 +标准雪纳瑞犬 +苏格兰梗犬 +西藏梗 +丝毛梗 +爱尔兰软毛梗犬 +西高地白梗 +拉萨阿普索犬 +平毛寻回犬 +卷毛寻回犬 +金毛猎犬 +拉布拉多猎犬 +乞沙比克猎犬 +德国短毛指示犬 +维兹拉犬 +英国塞特犬 +爱尔兰雪达犬 +戈登雪达犬 +布列塔尼犬猎犬 +小型西班牙猎犬 +英国史宾格犬 +威尔士史宾格犬 +可卡犬 +萨塞克斯猎犬 +爱尔兰水猎犬 +库瓦兹犬 +舒柏奇犬 +比利时牧羊犬 +马里努阿犬 +伯瑞犬 +凯尔皮犬 +可蒙犬 +老英国牧羊犬 +喜乐蒂牧羊犬 +牧羊犬 +边境牧羊犬 +法兰德斯牧牛狗 +罗特韦尔犬 +德国牧羊犬 +多伯曼犬 +迷你杜宾犬 +大瑞士山地犬 +伯恩山犬 +阿策尔山犬 +恩特尔布赫山犬 +拳师狗 +斗牛獒 +藏獒 +法国斗牛犬 +大丹犬 +圣伯纳德狗 +哈士奇 +阿拉斯加雪橇犬 +西伯利亚雪橇犬 +斑点狗 +猴犬 +巴仙吉犬 +巴哥犬 +兰波格犬 +纽芬兰犬 +大白熊犬 +萨摩耶犬 +博美犬 +松狮 +毛狮犬 +布鲁塞尔格林芬犬 +彭布洛克威尔士科基犬 +卡狄根威尔斯柯基犬 +玩具贵宾犬 +迷你贵宾犬 +标准贵宾犬 +墨西哥无毛犬 +灰狼 +阿拉斯加苔原狼 +红狼 +郊狼 +澳洲野狗 +豺 +非洲野犬 +鬣狗 +红狐 +敏狐 +北极狐 +灰狐 +虎斑猫 +孟加拉猫 +波斯猫 +暹罗猫 +埃及猫 +美洲狮 +猞猁 +美洲豹 +雪豹 +捷豹 +狮子 +老虎 +猎豹 +棕熊 +美洲黑熊 +北极熊 +懒熊 +侏獴 +狐獴 +虎甲 +瓢虫 +步甲 +天牛 +金花虫 +蜣螂 +犀金龟 +象鼻虫 +苍蝇 +蜜蜂 +蚂蚁 +草蜢 +蟋蟀 +竹节虫 +蟑螂 +螳螂 +蝉 +叶蝉 +草蜻蛉 +蜻蜓 +豆娘 +优红蛱蝶 +小环蝴蝶 +帝王蝴蝶 +菜粉蝶 +粉蝶 +灰蝶 +海星 +海胆 +海参 +棉尾兔 +野兔 +安哥拉兔 +仓鼠 +刺猬 +狐松鼠 +土拨鼠 +海狸 +豚鼠 +栗色马 +斑马 +猪 +野猪 +疣猪 +河马 +公牛 +水牛 +美洲野牛 +公羊 +大角羊 +阿尔卑斯羱羊 +披紅狷羚 +高角羚 +瞪羚 +阿拉伯单峰骆驼 +大羊驼 +黄鼠狼 +水貂 +欧洲臭鼬 +黑足鼬 +水獭 +臭鼬 +獾 +犰狳 +三趾树懒 +红毛猩猩 +大猩猩 +黑猩猩 +长臂猿 +合趾猿长臂猿 +长尾猴 +赤猴 +狒狒 +猕猴 +黑叶猴 +东非黑白疣猴 +长鼻猴 +狨猴 +白头卷尾猴 +吼猴 +伶猴 +蜘蛛猴 +松鼠猴 +马达加斯加环尾狐猴 +大狐猴 +亚洲象 +非洲草原象 +小熊猫 +大熊猫 +杖鱼 +鳗鱼 +银鲑 +石美人 +小丑鱼 +鲟鱼 +雀鳝 +狮子鱼 +河豚 +算盘 +长袍 +学位袍 +手风琴 +原声吉他 +航空母舰 +客机 +飞艇 +祭坛 +救护车 +水陆两用车 +模拟时钟 +蜂房 +围裙 +垃圾桶 +攻击步枪 +背包 +面包店 +平衡木 +气球 +圆珠笔 +创可贴 +班卓琴 +栏杆 +杠铃 +理发椅 +理发店 +谷仓 +晴雨表 +桶 +手推车 +棒球 +篮球 +婴儿床 +巴松管 +游泳帽 +沐浴毛巾 +浴缸 +旅行车 +灯塔 +烧杯 +军帽 +啤酒瓶 +啤酒杯 +钟塔 +口水兜 +双人骑自行车 +比基尼 +活页夹 +双筒望远镜 +鸟舍 +船库 +双人雪橇 +饰扣式领带 +阔边女帽 +书橱 +书店 +瓶盖 +弓箭 +蝴蝶结领结 +铜制牌位 +胸罩 +防波堤 +铠甲 +扫帚 +水桶 +搭扣 +防弹背心 +高铁 +肉铺 +出租车 +大锅 +蜡烛 +大炮 +独木舟 +开瓶器 +开衫 +车镜 +旋转木马 +工具箱 +纸箱 +车轮 +取款机 +录音带 +卡带播放器 +城堡 +双体船 +CD播放器 +大提琴 +手机 +铁链 +铁丝网 +锁子甲 +链锯 +储物柜 +小衣橱 +风铃 +橱柜 +圣诞袜 +教堂 +电影院 +切肉刀 +悬崖屋 +斗篷 +木屐 +鸡尾酒调酒器 +咖啡杯 +咖啡机 +线圈 +组合锁 +电脑键盘 +糖果店 +集装箱船 +敞篷车 +瓶塞钻 +短号 +牛仔靴 +牛仔帽 +摇篮 +起重机 +安全帽 +板条箱 +小儿床 +砂锅 +槌球 +拐杖 +胸甲 +大坝 +书桌 +台式电脑 +有线电话 +尿布 +数字时钟 +数字手表 +餐桌 +抹布 +洗碗机 +盘式制动器 +码头 +狗拉雪橇 +圆顶 +门垫 +钻井平台 +鼓 +鼓槌 +哑铃 +荷兰锅 +电风扇 +电吉他 +电力机车 +组合电视柜 +信封 +浓缩咖啡机 +扑面粉 +女用长围巾 +文件柜 +消防船 +消防车 +火炉栏 +旗杆 +长笛 +折叠椅 +橄榄球头盔 +叉车 +喷泉 +钢笔 +四帷柱床 +运货车厢 +圆号 +煎锅 +毛皮大衣 +垃圾车 +防毒面具 +汽油泵 +高脚杯 +卡丁车 +高尔夫球 +高尔夫球车 +缆车 +锣 +礼服 +钢琴 +温室 +散热器格栅 +杂货店 +断头台 +小发夹 +头发喷雾 +半履带装甲车 +锤子 +带盖的长方形大容器 +吹风机 +手提电脑 +手帕 +硬盘 +口琴 +竖琴 +联合收割机 +斧头 +手枪皮套 +家庭影院 +蜂窝 +钩子 +有裙撑的裙子 +单杠 +马车 +沙漏 +iPod +熨斗 +南瓜灯笼 +牛仔裤 +吉普车 +T恤衫 +拼图 +人力车 +操纵杆 +和服 +护膝 +蝴蝶结 +实验室大褂 +长柄勺 +灯罩 +笔记本电脑 +割草机 +镜头盖 +开信刀 +图书馆 +救生艇 +点火器 +豪华轿车 +远洋班轮 +唇膏 +平底便鞋 +洗剂 +扬声器 +放大镜 +锯木厂 +磁罗盘 +邮袋 +信箱 +女游泳衣 +有肩带浴衣 +窨井盖 +沙球（一种打击乐器） +马林巴木琴 +面具 +火柴 +花柱 +迷宫 +量杯 +药箱 +巨石 +麦克风 +微波炉 +军装 +奶瓶 +迷你巴士 +迷你裙 +小型货车 +导弹 +连指手套 +搅拌钵 +房车 +福特T型车 +光猫 +修道院 +显示器 +电瓶车 +砂浆 +博士帽 +清真寺 +蚊帐 +踏板式摩托车 +山地自行车 +登山帐 +鼠标 +捕鼠器 +搬家货车 +动物的口套 +金属钉子 +颈托 +项链 +橡皮奶嘴 +平板电脑 +方尖碑 +双簧管 +小鹅笛 +里程表 +滤油器 +风琴 +示波器 +罩裙 +牛车 +氧气面罩 +包装 +船桨 +明轮 +挂锁 +画笔 +睡衣 +宫殿 +排箫 +纸巾 +降落伞 +双杠 +公园长椅 +停车收费表 +有轨电车 +露台 +付费电话 +基座 +铅笔盒 +卷笔刀 +香水 +培养皿 +复印机 +拨弦片 +尖顶头盔 +用尖板条连成的尖桩篱栅 +皮卡车 +桥墩 +存钱罐 +药瓶 +枕头 +乒乓球 +风车 +海盗船 +饮料罐 +木工刨 +天文馆 +塑料袋 +板架 +犁型铲雪机 +手压皮碗泵 +宝丽来相机 +电线杆 +警车 +雨披 +台球桌 +汽水瓶 +花盆 +陶工旋盘 +电钻 +祈祷垫 +打印机 +监狱 +导弹 +投影仪 +冰球 +沙包 +小钱袋 +羽管笔 +被子 +赛车 +球拍 +散热器 +收音机 +射电望远镜 +雨桶 +休闲车 +卷轴 +反射式照相机 +冰箱 +遥控器 +餐厅 +左轮手枪 +步枪 +摇椅 +电转烤肉架 +橡皮 +橄榄球 +直尺 +跑步鞋 +保险柜 +安全别针 +盐瓶 +凉鞋 +纱笼 +萨克斯管 +剑鞘 +秤 +校车 +帆船 +记分牌 +CRT显示器 +螺丝 +螺丝刀 +安全带 +缝纫机 +盾牌 +皮鞋店 +日式拉门 +购物篮 +购物车 +铁锹 +浴帽 +浴帘 +滑雪板 +滑雪面罩 +睡袋 +滑尺 +滑动门 +角子老虎机 +潜水通气管 +摩托雪橇 +扫雪机 +皂液器 +足球 +袜子 +太阳能集热器 +宽边帽 +汤碗 +空格键 +空间加热器 +航天飞机 +锅铲 +快艇 +蜘蛛网 +纺锤 +跑车 +聚光灯 +舞台 +蒸汽机车 +钢拱桥 +钢滚筒 +听诊器 +女用披肩 +石头墙 +秒表 +火炉 +过滤器 +有轨电车 +担架 +沙发床 +佛塔 +潜水艇 +套装 +日晷 +太阳镜 +太阳镜 +防晒霜 +悬索桥 +拖把 +运动衫 +游泳裤 +秋千 +开关 +注射管 +台灯 +坦克 +录音机 +茶壶 +泰迪熊 +电视 +网球 +茅草 +幕布 +顶针 +打谷机 +王座 +瓦屋顶 +烤面包机 +烟草店 +马桶 +火炬 +图腾柱 +拖车 +玩具店 +拖拉机 +半挂式卡车 +托盘 +风衣 +三轮车 +三体船 +三脚架 +凯旋门 +无轨电车 +长号 +浴盆 +旋转式栅门 +打字机键盘 +雨伞 +独轮车 +直立式钢琴 +吸尘器 +花瓶 +拱顶 +天鹅绒 +自动售货机 +祭衣 +高架桥 +小提琴 +排球 +华夫烤盘 +挂钟 +钱包 +衣柜衣橱 +军用飞机 +洗脸盆 +洗衣机 +水瓶 +水壶 +水塔 +威士忌壶 +哨子 +假发 +纱窗 +百叶窗 +温莎领带 +葡萄酒瓶 +机翼 +炒菜锅 +木勺子 +毛织品 +原木栅栏 +沉船 +双桅船 +蒙古包 +网站 +漫画 +纵横字谜 +路标 +交通信号灯 +防尘罩 +菜单 +盘子 +牛油果酱 +清炖肉汤 +火锅 +果酱松糕 +冰淇淋 +冰棍 +法棍 +百吉饼 +椒盐脆饼 +芝士汉堡 +热狗 +土豆泥 +卷心菜 +西兰花 +花椰菜 +西葫芦 +意面南瓜 +绿色小南瓜 +南瓜 +黄瓜 +洋蓟 +甜椒 +刺棘蓟 +蘑菇 +澳洲青苹 +草莓 +橘子 +柠檬 +无花果 +菠萝 +香蕉 +菠萝蜜 +番荔枝 +石榴 +干草 +培根蛋酱意大利面 +巧克力酱 +面团 +肉卷 +披萨 +馅饼 +卷饼 +红葡萄酒 +意式浓缩咖啡 +茶杯 +蛋酒 +高山 +泡泡 +悬崖 +珊瑚礁 +间歇泉 +湖边 +岬角 +沙洲 +沙滩 +峡谷 +火山 +棒球运动员 +新郎 +潜水员 +油菜 +雏菊 +黄色杓兰 +玉米 +橡子 +玫瑰果 +七叶树果实 +珊瑚菌 +木耳 +鹿花菌 +臭角菇 +地星 +野生舞茸 +牛肝菌 +玉米棒子 +卫生纸 diff --git a/deployment.md b/deployment.md index 6bfd856..e2f2d3f 100644 --- a/deployment.md +++ b/deployment.md @@ -189,19 +189,19 @@ Finished ONNX to TensorRT conversion... 模型规模TensorRT图像侧模型TensorRT文本侧模型 - CN-CLIPRN50DownloadDownload + CN-CLIPRN50🤗Download🤗Download - CN-CLIPViT-B/16DownloadDownload + CN-CLIPViT-B/16🤗Download🤗Download - CN-CLIPViT-L/14DownloadDownload + CN-CLIPViT-L/14🤗Download🤗Download - CN-CLIPViT-L/14@336pxDownloadDownload + CN-CLIPViT-L/14@336px🤗Download🤗Download - CN-CLIPViT-H/14DownloadDownload + CN-CLIPViT-H/14🤗Download🤗Download
diff --git a/deployment_En.md b/deployment_En.md index a5539e1..96013fd 100644 --- a/deployment_En.md +++ b/deployment_En.md @@ -191,19 +191,19 @@ After the execution of the above sample code, we obtained ViT-B-16 size Chinese- Model SizeTensorRT Image-side ModelTensorRT Text-side Model - CN-CLIPRN50DownloadDownload + CN-CLIPRN50DownloadDownload - CN-CLIPViT-B/16DownloadDownload + CN-CLIPViT-B/16DownloadDownload - CN-CLIPViT-L/14DownloadDownload + CN-CLIPViT-L/14DownloadDownload - CN-CLIPViT-L/14@336pxDownloadDownload + CN-CLIPViT-L/14@336pxDownloadDownload - CN-CLIPViT-H/14DownloadDownload + CN-CLIPViT-H/14DownloadDownload diff --git a/zeroshot_dataset.md b/zeroshot_dataset.md index ac6d1c3..a48e28d 100644 --- a/zeroshot_dataset.md +++ b/zeroshot_dataset.md @@ -4,9 +4,9 @@ 本数据集为[ELEVATER Benchmark](https://eval.ai/web/challenges/challenge-page/1832)的图像分类基准的**中文版**，共包括20个图像分类数据集，包括Caltech-101、CIFAR-10、CIFAR-100、MNIST等。我们提供整理好的数据集，可以直接接入Chinese CLIP的代码进行零样本分类。 -下载链接：[点击这里](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/ELEVATER_all.zip) +下载链接： (37GB) -ImageNet的原始数据请自行到官网下载（可参考[此文档](https://gist.github.com/antoinebrl/7d00d5cb6c95ef194c737392ef7e476a)下载并将验证集转为ImageFolder格式），本项目仅提供[中文标签](http://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/ImageNet-1K/label_cn.txt)和[英文标签](http://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/ImageNet-1K/label.txt)。 +ImageNet的原始数据请自行到官网下载（可参考[此文档](https://gist.github.com/antoinebrl/7d00d5cb6c95ef194c737392ef7e476a)下载并将验证集转为ImageFolder格式），本项目仅提供[中文标签](datasets/ImageNet-1K/label_cn.txt)和[英文标签](datasets/ImageNet-1K/label.txt)。 ## 数据集说明 diff --git a/zeroshot_dataset_en.md b/zeroshot_dataset_en.md index 05f7067..bc1331d 100644 --- a/zeroshot_dataset_en.md +++ b/zeroshot_dataset_en.md @@ -4,9 +4,9 @@ The collection of dataset is the Chinese version of the Image Classification in the Wild in the [ELEVATER Benchmark](https://eval.ai/web/challenges/challenge-page/1832). It consists of 20 datasets, including Caltech-101, CIFAR-10, CIFAR-100, MNIST, etc. We provide our organized datasets, which enable direct usage of our codes on the datasets. -Download link: [Click here](https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/ELEVATER_all.zip) +Download link: (37GB) -For the ImageNet data, please visit the official website ([link](http://image-net.org)). You can refer to [this doc](https://gist.github.com/antoinebrl/7d00d5cb6c95ef194c737392ef7e476a) to prepare the validation set into ImageFolder format. This project only provides the label names in [Chinese](http://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/ImageNet-1K/label_cn.txt) and [English](http://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/datasets/ImageNet-1K/label.txt). +For the ImageNet data, please visit the official website ([link](http://image-net.org)). You can refer to [this doc](https://gist.github.com/antoinebrl/7d00d5cb6c95ef194c737392ef7e476a) to prepare the validation set into ImageFolder format. This project only provides the label names in [Chinese](datasets/ImageNet-1K/label_cn.txt) and [English](datasets/ImageNet-1K/label.txt). ## Notes We have organized 20 datasets into 20 directories, and zipped and uploaded them. Users can click the link above to download all the datasets. After unzipping the `ELEVATER_all.zip`, you will get the zipped files of each dataset in ELEVATER. Once again after unzipping the dataset zipfile, you will get the dataset directory with the following folder structure: