{"schema_version":"onlylabs.public_analysis_evidence.v1","title":"ByteDance (Doubao/Seed) analysis evidence pack","description":"Public onlylabs evidence pack for cited agent analysis: captured pages, ranked public signals, and stored web-search provenance used by the background analysis workflow.","url":"https://onlylabs.fyi/analysis/bytedance","json_url":"https://onlylabs.fyi/analysis/bytedance/evidence.json","generated_at":"2026-06-11T18:07:29.684Z","org":{"slug":"bytedance","name":"ByteDance (Doubao/Seed)","category":"frontier-lab","category_label":"Frontier lab","dossier_url":"https://onlylabs.fyi/labs/bytedance"},"analysis":{"url":"https://onlylabs.fyi/analysis/bytedance","json_url":"https://onlylabs.fyi/analysis/bytedance/analysis.json","generated_at":"2026-06-08T15:59:08.419+00:00"},"workflow":{"version":"synthesize-analyses","provider":null,"model":null,"agent":null,"public_pack_mode":"local-pages-and-events","live_web_fetches":false,"note":"Public evidence exports do not trigger live Exa calls; stored Exa provenance is included when analysis metadata contains it."},"stats":{"pages":28,"events":99,"web":0,"evidence":88,"signal_desks":{"hiring":0,"forks":1,"releases":34,"talking":0,"repos":25},"data_radar_lanes":{"data":2,"evals":3,"infrastructure":3,"safety":0,"product":0},"data_radar_matches":7,"stored_analysis_evidence":null,"stored_analysis_web":null,"stored_analysis_signal_desks":null,"stored_analysis_data_radar_lanes":null,"stored_analysis_data_radar_matches":null},"stored_web_provenance":null,"evidence":[{"ref":"P1","kind":"page","title":"ByteDance-Seed/decoupleQ repository metadata","date":"2026-06-11T03:58:38.14532+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/decoupleQ","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/decoupleQ\n\nDescription: A quantization algorithm for LLM\n\nLanguage: Cuda\n\nLicense: Apache-2.0\n\nStars: 151\n\nForks: 10\n\nOpen issues: 16\n\nCreated: 2024-04-19T08:18:27Z\n\nPushed: 2024-06-21T03:29:25Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# decoupleQ: Towards 2-bit Post-Training Uniform Quantization via decoupling Parameters into Integer and Floating Points\n\nThis repository contains the code for decoupleQ, the paper link is https://arxiv.org/abs/2404.12759 \n\nThe W2 CUDA kernel is available at https://github.com/NVIDIA/TensorRT-LLM/pull/1568\n\nSome of the code in this repo is built on top of [OPTQ's repository](https://github.com/IST-DASLab/gptq). We sincerely thank OPTQ for their great contribution.\n\nPlease feel free to raise issues or contact chenwei.gavin@bytedance.com or guoyi.0@bytedance.com if you have any question.\n\n## Dependencies\nAll of our experiments are conducted in the following environment.\n* datasets==1.17.0\n* transformers==4.35.0\n* torch==2.1.0\n\n## Reproduce\nTo reproduce the results of LLama, you should first download the models from [here](https://llama.meta.com/llama-downloads/), \nthen put it at ``MODEL_PATH``. Change the ``MODEL_PATH`` in the following command to the destination where the models are placed.\n```\nbash run_llama.sh MODEL_PATH # will get result 9.49 for wikiText2\nbash run_resnet.sh # will get result 64.134 for ResNet-18\n````\nIn llama quantization, if you find that the reproduced results (including the runtime) are far from the reported results, \nconsider modifying the flag: `torch.backends.cuda.matmul.allow_tf32`. More details can be found in [here](https://pytorch.org/docs/stable/notes/cuda.html#tf32-on-ampere).\n\nto run inference demo, you should first modify the ``build.sh``, change the ``DCMAKE_PREFIX_PATH``, ``DDECOUPLEQ_TORCH_HOME``, \n``DDECOUPLEQ_CUDA_HOME`` and ``DDECOUPLEQ_CUDNN_HOME`` based on your system, and then run the following commands:\n```\ngit submodule update --init\nbash build.sh # need cmake3.21+\nbash run_inference_llama.sh $LLAMA_ORG_MODEL_DIR $LLAMA_TRUE_QUANT_MODEL_PT\n```\n\n## Results\nHere is a summary of LLama results (runtime for\nthe quantization process is measured in hours):\n\n![dec"},{"ref":"P2","kind":"page","title":"ByteDance-Seed/ShadowKV repository metadata","date":"2026-06-11T03:58:38.031565+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/ShadowKV","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/ShadowKV\n\nDescription: [ICML 2025 Spotlight] ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 306\n\nForks: 23\n\nOpen issues: 8\n\nCreated: 2024-10-22T02:32:21Z\n\nPushed: 2025-05-01T22:49:30Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n<h1><img src=\"static/images/ShadowKV.png\" height=\"40px\"> ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference</h1>\n\n**training-free, high-throughput long-context LLM inference**\n</div>\n<div align=\"center\">\n<b><a href=\"https://github.com/preminstrel\">Hanshi Sun</a></b><sup>1,2</sup>,\n<b><a href=\"https://lchang20.github.io/\">Li-Wen Chang</a></b><sup>2</sup>,\n<b><a href=\"https://sites.google.com/view/wenleibao/\">Wenlei Bao</a></b><sup>2</sup>,\n<b><a href=\"https://sizezheng.github.io/\">Size Zheng</a></b><sup>2</sup>,\n<b><a href=\"https://zheng-ningxin.github.io/\">Ningxin Zheng</a></b><sup>2</sup>,\n<b><a href=\"https://scholar.google.com/citations?user=ZMfk2F8AAAAJ&hl=zh-CN\">Xin Liu</a></b><sup>2</sup>,\n<br>\n<b><a href=\"https://www.andrew.cmu.edu/user/harryd/\">Harry Dong</a></b><sup>1</sup>,\n<b><a href=\"https://users.ece.cmu.edu/~yuejiec/\">Yuejie Chi</a></b><sup>1</sup>,\n<b><a href=\"https://www.andrew.cmu.edu/user/beidic/\">Beidi Chen</a></b><sup>1</sup>\n</div>\n<div align=\"center\">\n<sup>1</sup>Carnegie Mellon University\n<sup>2</sup>ByteDance Seed\n</div>\n<div align=\"center\">\n[<a href=\"https://arxiv.org/abs/2410.21465\">Paper</a>] | [<a href=\"https://ByteDance-Seed.github.io/ShadowKV\">Blog</a>]\n</div>\n<br>\n\n<div align=\"center\">\n<img src=\"static/images/icml_frame.png\" align=\"top\"/>\n<figcaption>ShadowKV Framework</figcaption>\n</div>\n\n## 🔥 News\n- **[2025.05]** ShadowKV has been accepted by **ICML 2025** as **Spotlight**!\n- **[2024.10]** We have released the code.\n\n## Environment Set Up\nTo reproduce the results in the paper, you need to set up the environment as follows with a single A100 GPU:\n```bash\n# create env\nconda create -n ShadowKV python=3.10 -y\nconda activate ShadowKV\n\n# install packages\npip install -r requirements.txt\npip install flash-attn --no-build-isolation\n\n# nemo dependencies (for dataset "},{"ref":"P3","kind":"page","title":"ByteDance-Seed/FlexPrefill repository metadata","date":"2026-06-11T03:58:37.858559+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/FlexPrefill","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/FlexPrefill\n\nDescription: Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 169\n\nForks: 9\n\nOpen issues: 9\n\nCreated: 2025-02-18T07:02:28Z\n\nPushed: 2025-10-13T08:51:09Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<h1 align=\"center\">FlexPrefill</h1>\n\n<div align=\"center\">\n\n[![arxiv](https://img.shields.io/badge/arXiv-2502.20766-b31b1b.svg)](https://arxiv.org/abs/2502.20766)\n[![openreview](https://img.shields.io/badge/OpenReview-Paper-COLOR.svg)](https://openreview.net/forum?id=OfjIlbelrT)\n\n</div>\n\nThis repository provides the code for the paper [FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference](https://openreview.net/forum?id=OfjIlbelrT). \n\n**FlexPrefill** is selected as **Oral** Presentation(1.77%) at **ICLR 2025**!\n\n## TL;DR\n\nFlexPrefill is a dynamic and context-aware sparse attention mechanism that optimizes computational efficiency during long-sequence inference for large language models (LLMs). It achieves this by dynamically adjusting sparse attention patterns and computational budgets in real-time based on input demands and attention head requirements.\n\n## Requirements\n\nTo use FlexPrefill, you will need the following packages:\n\n- `torch==2.4.0`\n- `triton==3.0.0`\n- `transformers==4.44.0`\n- `flash_attn==2.6.3` (optional)\n- `vllm==0.5.4` (optional)\n\n## Installation\n\nYou can install FlexPrefill using pip:\n```shell\npip install git+https://github.com/bytedance/FlexPrefill.git\n```\n\n## Quick Start\n\n### Example Test\n\nYou can execute the `tests/test_llm.py` script to run a basic test on a specified model. This test includes examples with token lengths ranging from 4k to 128k and logs the model's total execution time.\n\n```shell\n# default transformers model inference\npython tests/test_llm.py --model meta-llama/Llama-3.1-8B-Instruct --pattern default\n# sparse attention inference\npython tests/test_llm.py --model meta-llama/Llama-3.1-8B-Instruct --pattern flex_prefill\n```\n\n### FlexPrefill Sparse Attention Function\n\nYou can invoke flex prefill sparse attention using the following co"},{"ref":"P4","kind":"page","title":"ByteDance-Seed/SDP4Bit repository metadata","date":"2026-06-11T03:58:37.851396+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/SDP4Bit","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/SDP4Bit\n\nDescription: official implementation of paper SDP4Bit: Toward 4-bit Communication Quantization in Sharded Data Parallelism for LLM Training\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 44\n\nForks: 8\n\nOpen issues: 0\n\nCreated: 2024-11-21T06:24:38Z\n\nPushed: 2024-12-11T04:37:04Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# SDP4Bit\nThis repository is the official implement of paper **[SDP4Bit: Toward 4-bit Communication Quantization in Sharded Data Parallelism for LLM Training](https://arxiv.org/abs/2410.15526)**.\n\n## Overview\nSDP4Bit is a communication quantization strategy designed to reduce the overhead of large-scale distributed training in Sharded Data Parallelism (ShardedDP). By utilizing quantization on weight differences and two-level gradient smooth quantization, SDP4Bit reduces the communication of weights and gradients to nearly 4 bits without compromising accuracy. \n\n## Paper Results Reproduce\n### Preparing for Data\nIn the data processing step, we followed the [data preprocessing instructions](https://github.com/NVIDIA/Megatron-LM?tab=readme-ov-file#data-preprocessing) in Megatron-LM official repository. We use the [**pile deduplicated dataset**](https://huggingface.co/datasets/EleutherAI/the_pile_deduplicated) provided by huggingface as our training baseline. For the vocabulary and merges file, we used same as gpt2 model. \n**Download**\n```\nfrom datasets import load_dataset\ntrain_data = load_dataset('EleutherAI/the_pile_deduplicated', split='train', num_proc=16)\ntrain_data.to_json(os.path.join(save_path, dataset_output_name), lines=True)\nhf_hub_download(repo_id=\"gpt2\", filename=\"merges.txt\", local_dir=save_path)\nhf_hub_download(repo_id=\"gpt2\", filename=\"vocab.json\", local_dir=save_path)\n```\n**Data Process**\nWe used [preprocess script](https://github.com/NVIDIA/Megatron-LM/blob/main/tools/preprocess_data.py) in Megatron-LM repository and the dataset download in last step. \n```\npython preprocess_data.py \\\n--input pile.jsonl \\\n--split train \\\n--columns text \\\n--output-prefix pile \\\n--vocab-file vocab.json \\\n--merge-file merges.txt \\\n--dataset-impl mmap \\\n--tokenizer-type GPT2BPETokenizer \\\n--append-eod \\\n--torch-backend"},{"ref":"P5","kind":"page","title":"ByteDance-Seed/VideoWorld repository metadata","date":"2026-06-11T03:58:37.536927+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/VideoWorld","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/VideoWorld\n\nDescription: [CVPR 2025] VideoWorld is a simple generative model that learns purely from unlabeled videos—much like how babies learn by observing their environment.\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 790\n\nForks: 41\n\nOpen issues: 13\n\nCreated: 2025-01-15T06:49:23Z\n\nPushed: 2026-02-25T11:23:41Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# VideoWorld Series: Learning World Models from Unlabeled Videos\n\nThis repository hosts the **VideoWorld** research project series, exploring how deep generative models can learn complex world knowledge, physics, and dynamics solely from visual inputs.\n\nThis project encompasses two main iterations:\n- **[VideoWorld (CVPR 2025)](./VideoWorld)**: The first generation model using Latent Dynamics Model (LDM) for knowledge acquisition.\n- **[VideoWorld 2 (CVPR 2026)](./VideoWorld2)**: The second generation focusing on *transferable* knowledge using disentangled Latent Dynamics Model (dLDM).\n\n---\n\n## 📂 Projects Overview\n\n### 1. [VideoWorld: Exploring Knowledge Learning from Unlabeled Videos](./VideoWorld)\n**Accepted by CVPR 2025**\n\n> **Highlights:**\n> * demonstrates that video generation models can learn complex rules (e.g., Go game) without reward signals.\n> * Introduces the **Latent Dynamics Model (LDM)** to compress visual changes into informative latent codes.\n> * Achieves 5-dan professional level in Go and strong performance in robotic control tasks (CALVIN).\n\n* **Code:** [./VideoWorld](./VideoWorld)\n* **Paper:** [arXiv](https://arxiv.org/pdf/2501.09781)\n\n### 2. [VideoWorld 2: Learning Transferable Knowledge from Real-world Video](./VideoWorld2)\n\n> **Highlights:**\n> * Focuses on **transferable knowledge** and long-horizon tasks in real-world settings.\n> * Proposes the **disentangled Latent Dynamics Model (dLDM)** to decouple action dynamics from visual appearance.\n> * Significant improvements in task success rates (up to 70%) on challenging handcraft benchmarks.\n\n* **Code:** [./VideoWorld2](./VideoWorld2)\n\n---\n\n## 🚀 Getting Started\n\nSince the two projects were developed with slightly different dependencies to maintain reproducibility, we recommend using separate environments for ea"},{"ref":"P6","kind":"page","title":"ByteDance-Seed/Agent-R repository metadata","date":"2026-06-11T03:58:37.413145+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Agent-R","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Agent-R\n\nDescription: Resources for our paper: \"Agent-R: Training Language Model Agents to Reflect via Iterative Self-Training\"\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 172\n\nForks: 20\n\nOpen issues: 0\n\nCreated: 2025-01-15T10:51:25Z\n\nPushed: 2025-10-20T02:30:17Z\n\nDefault branch: main\n\nFork: no\n\nArchived: yes\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone! \n<br>\nWe are <b>ByteDance Seed team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://team.doubao.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n<a href=\"https://github.com/user-attachments/assets/93481cda-a7f3-47f3-b333-fe6b3da86b78\">\n<img src=\"https://img.shields.io/badge/WeChat-07C160?style=for-the-badge&logo=wechat&logoColor=white\"></a>\n<a href=\"https://www.xiaohongshu.com/user/profile/668e7e15000000000303157d?xsec_token=ABl2-aqekpytY6A8TuxjrwnZskU-6BsMRE_ufQQaSAvjc%3D&xsec_source=pc_search\">\n<img src=\"https://img.shields.io/badge/Xiaohongshu-%23FF2442?style=for-the-badge&logo=xiaohongshu&logoColor=white\"></a>\n<a href=\"https://www.zhihu.com/org/dou-bao-da-mo-xing-tuan-dui/\">\n<img src=\"https://img.shields.io/badge/zhihu-%230084FF?style=for-the-badge&logo=zhihu&logoColor=white\"></a>\n</p>\n\n![seed logo](https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216)\n\n<h1 align=\"center\">Agent-R: Training Language Model Agents to Reflect via Iterative Self-Training</h1>\n<p align=\"center\">\n<a href=\"https://arxiv.org/abs/2501.11425\">\n<img src=\"https://img.shields.io/badge/AgentR-Paper-red\"></a>\n<a href=\"https://github.com/bytedance/Agent-R\">\n<img src=\"https://img.shields.io/badge/AgentR-Project Page-yellow\"></a>\n</p>\n\n# Updates\n+ [2025.01.21] We release Agent-R.\n+ 🔥 The paper is available at [Agent-R Paper](https://arxiv.org/abs/2501.11425).\n+ 🔥 The code is available at [Agent-R Code](https://github.com/bytedance/Agent-R).\n\n# Introduction\nWe propose an iterative self-training framework, **Agent-R**, that enables language Agent to Reflect on the fly. Unlike traditional methods that reward or penalize actions solely based on correctness, ou"},{"ref":"P7","kind":"page","title":"ByteDance-Seed/ByteCheckpoint repository metadata","date":"2026-06-11T03:58:37.167815+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/ByteCheckpoint","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/ByteCheckpoint\n\nDescription: ByteCheckpoint: An Unified Checkpointing Library for LFMs\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 282\n\nForks: 19\n\nOpen issues: 3\n\nCreated: 2025-03-20T04:03:05Z\n\nPushed: 2026-02-02T06:01:33Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone! \n<br>\nWe are <b>ByteDance Seed team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://team.doubao.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n<a href=\"https://github.com/user-attachments/assets/93481cda-a7f3-47f3-b333-fe6b3da86b78\">\n<img src=\"https://img.shields.io/badge/WeChat-07C160?style=for-the-badge&logo=wechat&logoColor=white\"></a>\n<a href=\"https://www.xiaohongshu.com/user/profile/668e7e15000000000303157d?xsec_token=ABl2-aqekpytY6A8TuxjrwnZskU-6BsMRE_ufQQaSAvjc%3D&xsec_source=pc_search\">\n<img src=\"https://img.shields.io/badge/Xiaohongshu-%23FF2442?style=for-the-badge&logo=xiaohongshu&logoColor=white\"></a>\n<a href=\"https://www.zhihu.com/org/dou-bao-da-mo-xing-tuan-dui/\">\n<img src=\"https://img.shields.io/badge/zhihu-%230084FF?style=for-the-badge&logo=zhihu&logoColor=white\"></a>\n</p>\n\n![seed logo](https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216)\n\n# ByteCheckpoint: A Unified Checkpointing System for Large Foundation Model Development\n<p align=\"center\">\n<a href=\"https://arxiv.org/pdf/2407.20143\">\n<img src=\"https://img.shields.io/badge/Paper-NSDI-red\"></a>\n<a href=\"LICENSE\">\n<img src=\"https://img.shields.io/badge/License-Apache-blue\"></a>\n</p>\n\nByteCheckpoint is a unified, efficient and production-grade checkpointing system for large foundation model development.\n\nByteCheckpoint is the open-source implementation of our research paper:\n[ByteCheckpoint: A Unified Checkpointing System for Large Foundation Model Development](https://arxiv.org/abs/2407.20143).\n\nByteCheckpoint is easy to use and efficient with:\n\n✔ **Framework-Agnostic API**: Provides a unified checkpointing entrypoint, i.e., `bytecheckpoint.save` and `bytecheckpoint.load`, to support various paralle"},{"ref":"P8","kind":"page","title":"ByteDance-Seed/VeOmni repository metadata","date":"2026-06-11T03:58:36.95602+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/VeOmni","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/VeOmni\n\nDescription: VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 2003\n\nForks: 211\n\nOpen issues: 95\n\nCreated: 2025-03-28T03:42:42Z\n\nPushed: 2026-06-10T02:20:41Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n\n<img src=\"./docs/assets/logo.png\" width=\"50%\">\n\n<div align=\"center\">\nVeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo\n<br>\n<br>\n</div>\n\n[![GitHub Repo stars](https://img.shields.io/github/stars/ByteDance-Seed/VeOmni)](https://github.com/ByteDance-Seed/VeOmni/stargazers)\n[![Paper](https://img.shields.io/badge/Paper-red)](https://arxiv.org/abs/2508.02317)\n[![Documentation](https://img.shields.io/badge/Documentation-blue)](https://veomni.readthedocs.io/en/latest/)\n[![WeChat](https://img.shields.io/badge/WeChat-green?logo=wechat&amp)](https://raw.githubusercontent.com/ByteDance-Seed/VeOmni/refs/heads/main/docs/assets/wechat.png)\n\n</div>\n\n## 🍪 Overview\nVeOmni is a versatile framework for both single- and multi-modal pre-training and post-training. It empowers users to seamlessly scale models of any modality across various accelerators, offering both flexibility and user-friendliness.\n\nOur guiding principles when building VeOmni are:\n- **Flexibility and Modularity**: VeOmni is built with a modular design, allowing users to decouple most components and replace them with their own implementations as needed.\n- **Trainer-free**: VeOmni supports linear training scripts that avoid rigid, structured trainer classes (e.g., [PyTorch-Lightning](https://github.com/Lightning-AI/pytorch-lightning) or [HuggingFace](https://huggingface.co/docs/transformers/v4.50.0/en/main_classes/trainer#transformers.Trainer) Trainer). These training scripts expose the entire training logic to users for maximum transparency and control. Besides, VeOmni supports a basic trainer for text-only or vlm/omni models training and a rl trainer as a trainer backend in reinforcement learning.\n\n- **Omni model native**: VeOmni enables users to effortlessly scale any omni-model across devices and accelerators.\n- **Torch native**: VeOmni is"},{"ref":"P9","kind":"page","title":"ByteDance-Seed/Triton-distributed repository metadata","date":"2026-06-11T03:58:36.718027+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Triton-distributed","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Triton-distributed\n\nDescription: Distributed Compiler based on Triton for Parallel Systems\n\nLanguage: Python\n\nLicense: MIT\n\nStars: 1457\n\nForks: 150\n\nOpen issues: 46\n\nCreated: 2025-04-02T06:57:03Z\n\nPushed: 2026-04-22T09:57:16Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone!\n<br>\nWe are <b>ByteDance Seed team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://team.doubao.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n<a href=\"https://github.com/user-attachments/assets/93481cda-a7f3-47f3-b333-fe6b3da86b78\">\n<img src=\"https://img.shields.io/badge/WeChat-07C160?style=for-the-badge&logo=wechat&logoColor=white\"></a>\n<a href=\"https://www.xiaohongshu.com/user/profile/668e7e15000000000303157d?xsec_token=ABl2-aqekpytY6A8TuxjrwnZskU-6BsMRE_ufQQaSAvjc%3D&xsec_source=pc_search\">\n<img src=\"https://img.shields.io/badge/Xiaohongshu-%23FF2442?style=for-the-badge&logo=xiaohongshu&logoColor=white\"></a>\n<a href=\"https://www.zhihu.com/org/dou-bao-da-mo-xing-tuan-dui/\">\n<img src=\"https://img.shields.io/badge/zhihu-%230084FF?style=for-the-badge&logo=zhihu&logoColor=white\"></a>\n</p>\n\n![seed logo](https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216)\n\n# Triton-distributed\n\n<!-- <p align=\"center\">\n<a href=\"https://github.com/bytedance/flux\">\n<img src=\"https://img.shields.io/badge/Triton-distributed-Project Page-yellow\"></a>\n<a href=\"https://arxiv.org/pdf/xxxx.xxxx\">\n<img src=\"https://img.shields.io/badge/Triton-distributed-Tech Report-red\"></a>\n<br>\n<a href=\"https://github.com/user-attachments/assets/d3fcb3bf-466b-4efe-8c3f-5f85258202ae\">\n<img src=\"https://img.shields.io/badge/Triton-distributed-Wechat Communication Group-07C160\"></a>\n<a href=\"XXX\">\n<img src=\"https://img.shields.io/badge/License-MIT-blue\"></a>\n</p> -->\n\n[Original Triton README](https://github.com/triton-lang/triton/blob/main/README.md) | [README in Chinese](README-cn.md)\n\nTriton-distributed is a distributed compiler designed for computation-communication overlapping, which is based on OpenAI Trito"},{"ref":"P10","kind":"page","title":"ByteDance-Seed/Seed-Thinking-v1.5 repository metadata","date":"2026-06-11T03:58:36.669461+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Seed-Thinking-v1.5","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Seed-Thinking-v1.5\n\nStars: 812\n\nForks: 18\n\nOpen issues: 10\n\nCreated: 2025-04-10T10:05:49Z\n\nPushed: 2025-06-09T12:08:26Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align='center'>\n<h1>Seed-Thinking-v1.5: Advancing Superb Reasoning Models with Reinforcement Learning</h1>\n<!-- TODO: Thread,Paper,Dataset,Weights-->\n<!-- [![Paper](https://img.shields.io/badge/paper-5f16a8?style=for-the-badge&logo=arxiv&logoColor=white)]() -->\n<!-- [![Blog](https://img.shields.io/badge/Blog-3858bf?style=for-the-badge&logo=homepage&logoColor=white)]() -->\n<!-- [![Dataset](https://img.shields.io/badge/API-4d8cd8?style=for-the-badge&logo=huggingface&logoColor=white)]() -->\n<!-- [![API](https://img.shields.io/badge/API-63cad3?style=for-the-badge&logo=huggingface&logoColor=white)]() -->\n<!-- [![Thread](https://img.shields.io/badge/Thread-91ded6?style=for-the-badge&logo=x&logoColor=white)]() -->\n</div>\n\n<!-- # Introduction -->\n\nWe introduce Seed-Thinking-v1.5, capable of reasoning through thinking before responding, resulting in improved performance on a wide range of benchmarks. Seed-Thinking-v1.5 achieves 86.7 on AIME 2024, 55.0 on Codeforces and 77.3 on GPQA, demonstrating excellent reasoning abilities in STEM and coding. Beyond reasoning tasks, the method demonstrates notable generalization across diverse domains. For instance, it surpasses DeepSeek R1 by 8% in win rate on non-reasoning tasks, indicating its broader applicability. Compared to other state-of-the-art reasoning models, Seed-Thinking-v1.5 is a Mixture-of-Experts (MoE) model with a relatively small size, featuring 20B activated and 200B total parameters. As part of our effort to assess generalized reasoning, we develop two internal benchmarks, BeyondAIME and Codeforces, both of which will be publicly released to support future research.\n\n![Model Performance](images/performance.png)\n\n# Technical Details\n\nFull technical details can be found in our [technical report](https://github.com/ByteDance-Seed/Seed-Thinking-v1.5/blob/main/seed-thinking-v1.5.pdf).\n\n# Full Results\n\nWe present the evaluation results across diverse tasks spanning mathematics, coding, science, and general knowledge domains.\n\n"},{"ref":"P11","kind":"page","title":"ByteDance-Seed/SAIL repository metadata","date":"2026-06-11T03:58:36.236118+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/SAIL","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/SAIL\n\nDescription: Implementation for \"The Scalability of Simplicity: Empirical Analysis of Vision-Language Learning with a Single Transformer\"\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 84\n\nForks: 3\n\nOpen issues: 2\n\nCreated: 2025-04-19T02:04:51Z\n\nPushed: 2025-10-29T01:24:18Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# The Scalability of Simplicity: Empirical Analysis of Vision-Language Learning with a Single Transformer (SAIL)\n<p align=\"center\">\n<a href=\"https://github.com/bytedance/SAIL\">\n<img src=\"https://img.shields.io/badge/SAIL-Project Page-yellow\"></a>\n<a href=\"https://arxiv.org/abs/2504.10462\">\n<img src=\"https://img.shields.io/badge/SAIL-Tech Report-red\"></a>\n<a href=\"https://huggingface.co/ByteDance-Seed/SAIL-7B\">\n<img src=\"https://img.shields.io/badge/SAIL-Hugging Face-orange\"></a>\n<a href=\"LICENSE\">\n<img src=\"https://img.shields.io/badge/License-Apache2.0-blue\"></a>\n</p>\n\nWe are extremely delighted to release SAIL, a **S**ingle tr**A**nsformer model for v**I**sion and **L**anguage. SAIL is a unified multimodal large language model (MLLM) that seamlessly integrates raw pixel encoding and language decoding within a single architecture. **​Without relying on pre-trained vision encoders**, SAIL achieves competitive performance across a wide range of vision-language tasks and demonstrates strong visual representation, rivaling state-of-the-art vision models in tasks like semantic segmentation.\n\n## Model & Micro Design\n<div align=\"center\">\n<img src=\"assets/sail_model.jpg\" alt=\"model\" style=\"height: 300; width: auto;\">\n</div>\n\n## An Overview of Comparison\n(A) Data scaling curve for Modular Multimodal Large Language Model (MLLM) and SAIL, our Single Transformer-based MLLM. As pretraining data increases, SAIL shows a sharper performance gain, demonstrating its superior data scalability.\n(B) Comparison to existing Single Transformer-based MLLMs, our SAIL pushes the performance boundaries on both vision tasks and vision-language tasks.\n<div align=\"center\">\n<img src=\"assets/perf_cmp.jpg\" alt=\"cmp\" style=\"height: 250; width: auto;\">\n</div>\n\n# News\n- [2025/06/26]🎉SAIL is accepted to ICCV 2025 (Highlight).\n- [2025/04/02]🔥We releas"},{"ref":"P12","kind":"page","title":"ByteDance-Seed/Bagel repository metadata","date":"2026-06-11T03:58:36.232408+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Bagel","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Bagel\n\nDescription: Open-source unified multimodal model\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 6000\n\nForks: 533\n\nOpen issues: 151\n\nCreated: 2025-04-17T06:54:07Z\n\nPushed: 2026-05-04T17:01:02Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<p align=\"center\">\n<img src=\"https://lf3-static.bytednsdoc.com/obj/eden-cn/nuhojubrps/banner.png\" alt=\"BAGEL\" width=\"480\"/>\n</p>\n\n<p align=\"center\">\n<a href=\"https://bagel-ai.org/\">\n<img\nsrc=\"https://img.shields.io/badge/BAGEL-Website-0A66C2?logo=safari&logoColor=white\"\nalt=\"BAGEL Website\"\n/>\n</a>\n<a href=\"https://arxiv.org/abs/2505.14683\">\n<img\nsrc=\"https://img.shields.io/badge/BAGEL-Paper-red?logo=arxiv&logoColor=red\"\nalt=\"BAGEL Paper on arXiv\"\n/>\n</a>\n<a href=\"https://huggingface.co/ByteDance-Seed/BAGEL-7B-MoT\">\n<img \nsrc=\"https://img.shields.io/badge/BAGEL-Model-yellow?logo=huggingface&logoColor=yellow\" \nalt=\"BAGEL Model\"\n/>\n</a>\n<a href=\"https://demo.bagel-ai.org/\">\n<img\nsrc=\"https://img.shields.io/badge/BAGEL-Demo-blue?logo=googleplay&logoColor=blue\"\nalt=\"BAGEL Demo\"\n/>\n</a>\n<a href=\"https://huggingface.co/spaces/ByteDance-Seed/BAGEL\">\n<img \nsrc=\"https://img.shields.io/badge/BAGEL-Space-orange?logo=huggingface&logoColor=yellow\" \nalt=\"BAGEL Model\"\n/>\n</a>\n<a href=\"https://discord.gg/eXQNFhWe\">\n<img\nsrc=\"https://img.shields.io/badge/BAGEL-Discord-5865F2?logo=discord&logoColor=purple\"\nalt=\"BAGEL Discord\"\n/>\n</a>\n<a href=\"mailto:bagel@bytedance.com\">\n<img\nsrc=\"https://img.shields.io/badge/BAGEL-Email-D14836?logo=gmail&logoColor=red\"\nalt=\"BAGEL Email\"\n/>\n</a>\n</p>\n\n# Unified Model for Multimodal Understanding and Generation\n> [Chaorui Deng*](https://scholar.google.com/citations?hl=en&user=k0TWfBoAAAAJ), [Deyao Zhu*](https://tsutikgiau.github.io/), [Kunchang Li*](https://andy1621.github.io/), [Chenhui Gou*](https://www.linkedin.com/in/chenhui-gou-9201081a1/?originalSubdomain=au), [Feng Li*](https://fengli-ust.github.io/), [Zeyu Wang](https://zw615.github.io/), Shu Zhong, [Weihao Yu](https://whyu.me/), [Xiaonan Nie](https://codecaution.github.io/), [Ziang Song](https://www.linkedin.com/in/ziang-song-43b0ab8a/), Guang Shi :email: , [Haoqi Fan* :tophat: ](https://haoqifan.github.io/)\n>\n> contact: s"},{"ref":"P13","kind":"page","title":"ByteDance-Seed/Seed-Coder repository metadata","date":"2026-06-11T03:58:36.21493+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Seed-Coder","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Seed-Coder\n\nDescription: Seed-Coder is a family of lightweight open-source code LLMs comprising base, instruct and reasoning models, developed by ByteDance Seed.\n\nLicense: MIT\n\nStars: 754\n\nForks: 59\n\nOpen issues: 18\n\nCreated: 2025-04-21T06:52:36Z\n\nPushed: 2025-06-06T02:10:41Z\n\nDefault branch: master\n\nFork: no\n\nArchived: no\n\nREADME:\n# <img src=\"./imgs/logo.png\" height=\"25\"> Seed-Coder: Let the Code Model Curate Data for Itself\n\n<p align=\"center\">\n🌐 <a href=\"https://bytedance-seed-coder.github.io/\"> Homepage</a>&nbsp&nbsp | &nbsp&nbsp🤗 <a href=\"https://huggingface.co/collections/ByteDance-Seed/seed-coder-680de32c15ead6555c75b0e4\">Hugging Face</a>&nbsp&nbsp | &nbsp&nbsp📄 <a href=\"https://arxiv.org/abs/2506.03524\">arXiv</a>\n</p>\n\nWe are thrilled to introduce **Seed-Coder** (previously known as Doubao-Coder), a family of lightweight yet powerful open-source code LLMs comprising base, instruct and reasoning models of 8B size. \n\nSeed-Coder demonstrates that, with minimal human effort, LLMs can effectively curate code training data by themselves to drastically enhance coding capabilities.\n\nSeed-Coder represents our initial step towards contributing to the open-source LLM ecosystem. We look forward to seeing Seed-Coder drive advances in code intelligence and empower broader applications in the open-source LLM community!\n\n## 📢 News\n\n[2025/05/08]🔥 We release [Seed-Coder](https://bytedance-seed-coder.github.io)!\n\n[2025/05/14] We found an inconsistent setting of our BigCodeBench evaluation results. We already updated all the results with the aligned BigCodeBench-Completion setting.\n\n## 🌟 Highlights\n- **Model-centric:** Seed-Coder predominantly leverages LLMs instead of hand-crafted rules for code data filtering, minimizing manual effort in pretraining data construction.\n\n- **Transparent:** We openly share detailed insights into our model-centric data pipeline, including methods for curating GitHub data, commits data, and code-related web data.\n\n- **Powerful:** Seed-Coder achieves state-of-the-art performance among open-source models of comparable size across a diverse range of coding tasks.\n\n## ⚡ Quick Start\nWe are excited to introduce **Seed-Coder**, "},{"ref":"P14","kind":"page","title":"ByteDance-Seed/Chain-of-Action repository metadata","date":"2026-06-11T03:58:35.483163+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Chain-of-Action","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Chain-of-Action\n\nDescription: Official implementation of Chain-of-Action: Trajectory Autoregressive Modeling for Robotic Manipulation. Accepted in NeurIPS 2025.\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 104\n\nForks: 8\n\nOpen issues: 2\n\nCreated: 2025-07-03T06:19:29Z\n\nPushed: 2025-12-13T06:38:12Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<p align=\"center\">\n<img src=\"assets/logo.png\" alt=\"Project Logo\" width=\"640\"/>\n</p>\n\n<p align=\"center\">\n<a href=\"https://chain-of-action.github.io/\"><img src=\"https://img.shields.io/badge/Website-Visit-0A66C2?logo=safari&logoColor=white\" alt=\"Website\" /></a> <a href=\"https://arxiv.org/pdf/2506.09990\"><img src=\"https://img.shields.io/badge/Paper-arXiv-red?logo=arxiv&logoColor=red\" alt=\"Paper on arXiv\" /></a> <a href=\"https://huggingface.co/Solomonz/Chain-of-Action\"><img src=\"https://img.shields.io/badge/HuggingFace-Model-yellow?logo=huggingface&logoColor=yellow\" alt=\"HuggingFace Model\" /></a> <a href=\"https://huggingface.co/datasets/Solomonz/Chain-of-Action\"><img src=\"https://img.shields.io/badge/HuggingFace-Data-blue?logo=huggingface&logoColor=blue\" alt=\"HuggingFace Dataset\" /></a>\n</p>\n\n## Quick start\n### Set up environment\n\n```bash\nconda create -n coa python=3.9 -y\nconda activate coa\nbash scripts/init.sh\nsource ~/.bashrc\n```\n\ninstall dependencies and RLBench enviroment, see [init.sh](scripts/init.sh) for details\n\n### One-click Evaluation\n\nThe script will automatically download the required pretrained snapshot and the necessary evaluation dataset for the specified task. \n\n```bash\nbash scripts/eval.sh task=push_button\n```\n\n## Experiments Results\n\n### Evluation over 60 RLBench tasks\nWhy we use 60 tasks for the main evaluation?\nAlthough the 18 RLBench tasks have been widely adopted as a benchmark since their introduction in “Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation”, they are primarily used to evaluate 3D-based hierarchical policies that depend heavily on high-precision 3D inputs and motion planners. Many of these tasks are extremely challenging for RGB-only visuomotor policies, often leading to uniformly low success rates and therefore limited discriminative power.\n\n<img wi"},{"ref":"P15","kind":"page","title":"ByteDance-Seed/StragglerAnalysis repository metadata","date":"2026-06-11T03:58:35.346052+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/StragglerAnalysis","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/StragglerAnalysis\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 54\n\nForks: 4\n\nOpen issues: 0\n\nCreated: 2025-04-25T08:15:01Z\n\nPushed: 2025-04-30T03:58:52Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone!\n<br>\nWe are <b>ByteDance Seed team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://team.doubao.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n<a href=\"https://github.com/user-attachments/assets/469535a8-42f2-4797-acdf-4f7a1d4a0c3e\">\n<img src=\"https://img.shields.io/badge/WeChat-07C160?style=for-the-badge&logo=wechat&logoColor=white\"></a>\n<a href=\"豆包研究员 - 小红书\">\n<img src=\"https://img.shields.io/badge/Xiaohongshu-%23FF2442?style=for-the-badge&logo=xiaohongshu&logoColor=white\"></a>\n<a href=\"https://www.zhihu.com/org/dou-bao-da-mo-xing-tuan-dui/\">\n<img src=\"https://img.shields.io/badge/zhihu-%230084FF?style=for-the-badge&logo=zhihu&logoColor=white\"></a>\n</p>\n\n![seed logo](https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216)\n\n# Artifact for Paper \"Understanding Stragglers in Large Model Training Using What-if Analysis\"\n<p align=\"center\">\n<a href=\"XXX\">\n<img src=\"https://img.shields.io/badge/License-Apache2.0-blue\"></a>\n</p>\n\n## Introduction\nThis artifact provides the core functionality of the simulator and the what-if analysis proposed in the paper, along with three sample traces to demonstrate the usage of the tool. The expected output includes the following for each sample trace:\n- Estimated slowdown $S$ (i.e., Eq. 1)\n- Slowdown $S_t$ attributed to each operation type $t$ (i.e., Eq. 2)\n- Slowdown $S_w$ attributed to each worker $w$ (i.e., Eq. 4)\n- Characterization metrics $M_W$(i.e., Eq. 5) and $M_S$ for individual worker issues and stage partitioning imbalance, respectively\n- A heatmap visualization as in Fig. 14.\n- A timeline of the simulated ideal trace visualizable in Perfetto.\n\n## Code Structure\n```bash\n├── analyzer # Analyzer codes\n├── data # Stores input data for analysis and corresponding expected results\n├── format.sh # Script fo"},{"ref":"P16","kind":"page","title":"ByteDance-Seed/DeepFlow repository metadata","date":"2026-06-11T03:58:34.871073+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/DeepFlow","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/DeepFlow\n\nDescription: [ICCV 2025] Deeply Supervised Flow-Based Generative Models\n\nLanguage: Python\n\nLicense: NOASSERTION\n\nStars: 36\n\nForks: 2\n\nOpen issues: 0\n\nCreated: 2025-04-27T11:32:15Z\n\nPushed: 2025-06-26T03:13:42Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone! \n<br>\nWe are <b>ByteDance Seed team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://team.doubao.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n<a href=\"https://github.com/user-attachments/assets/93481cda-a7f3-47f3-b333-fe6b3da86b78\">\n<img src=\"https://img.shields.io/badge/WeChat-07C160?style=for-the-badge&logo=wechat&logoColor=white\"></a>\n<a href=\"https://www.xiaohongshu.com/user/profile/668e7e15000000000303157d?xsec_token=ABl2-aqekpytY6A8TuxjrwnZskU-6BsMRE_ufQQaSAvjc%3D&xsec_source=pc_search\">\n<img src=\"https://img.shields.io/badge/Xiaohongshu-%23FF2442?style=for-the-badge&logo=xiaohongshu&logoColor=white\"></a>\n<a href=\"https://www.zhihu.com/org/dou-bao-da-mo-xing-tuan-dui/\">\n<img src=\"https://img.shields.io/badge/zhihu-%230084FF?style=for-the-badge&logo=zhihu&logoColor=white\"></a>\n</p>\n\n![seed logo](https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216)\n\n<!-- 注释：以上为Seed官方信息，可直接复制使用，请注意导入“Seed WeChat”（第12行）、“Seed logo”(第20行)图片替换 -->\n\n# Deeply Supervised Flow-Based Generative Models\n\n### [ICCV 2025]\n\n<p align=\"center\">\n<a href=\"https://deepflow-project.github.io/\">\n<img src=\"https://img.shields.io/badge/DeepFlow-Project Page-yellow\"></a>\n<a href=\"https://arxiv.org/abs/2503.14494\">\n<img src=\"https://img.shields.io/badge/DeepFlow-Tech Report-red\"></a>\n<!-- <a href=\"https://www.apache.org/licenses/\">\n<img src=\"https://img.shields.io/badge/License-Apache 2.0-blue\"></a> -->\n</p>\n\n<p align=\"center\">\n<a href=\"https://dlsrbgg33.github.io/\" target=\"_blank\">Inkyu&nbsp;Shin</a> &ensp; <b>&middot;</b> &ensp;\n<a href=\"https://www.chenglinyang.com/\" target=\"_blank\">Chenglin&nbsp;Yang</a> &ensp; <b>&middot;</b> &ensp;\n<a href=\"http://liangchiehchen.com/\" target=\"_blank\">Liang-Chieh&nbsp;"},{"ref":"P17","kind":"page","title":"ByteDance-Seed/.github repository metadata","date":"2026-06-11T03:58:34.64475+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/.github","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/.github\n\nStars: 2\n\nForks: 2\n\nOpen issues: 0\n\nCreated: 2025-04-23T09:57:10Z\n\nPushed: 2025-04-29T04:43:37Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Bytedance-Seed\n## 简介\n字节跳动 Seed 团队成立于 2023 年，致力于寻找通用智能的新方法，追求智能上限。团队研究方向涵盖 LLM、语音、视觉、世界模型、基础架构、AI Infra、下一代 AI 交互等，在中国、新加坡、美国等地设有实验室和岗位。\nSeed 团队在 AI 领域拥有长期愿景与决心，坚持深耕基础，期望成为世界一流的 AI 研究团队，为科技和社会发展作出贡献。目前团队已推出业界领先的通用大模型以及前沿的多模态能力，支持豆包、扣子、即梦等超过 50 个应用场景。\n\nEstablished in 2023, the ByteDance Seed team is dedicated to discovering new approaches to general intelligence and pushing the boundaries of AI. Our research spans large language models, speech, vision, world models, AI infrastructure, next-generation interfaces and more. \nWith a long-term vision and determination in AI, the ByteDance Seed team remains committed to foundational research. We aim to become a world-class AI research team that drives real technological progress and delivers societal benefits.\nWith labs across China, Singapore, and the U.S., our team has already released industry-leading general-purpose large models and advanced multimodal capabilities, powering over 50 real-world applications — including Doubao, Coze, and Jimeng.\n## 网站链接\n[官网链接](https://seed.bytedance.com/)\n## 联系方式\n邮箱: doubao-llm@bytedance.com"},{"ref":"P18","kind":"page","title":"ByteDance-Seed/EvaLearn repository metadata","date":"2026-06-11T03:58:34.518613+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/EvaLearn","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/EvaLearn\n\nDescription: EvaLearn is a pioneering benchmark designed to evaluate large language models (LLMs) on their learning capability and efficiency in challenging tasks.\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 431\n\nForks: 12\n\nOpen issues: 0\n\nCreated: 2025-06-03T04:41:27Z\n\nPushed: 2026-05-12T07:00:26Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n<img src=\"logo.png\" alt=\"Bytedance-seed\" width=\"300\"/>\n</div>\n\n<div align=\"center\">\n<h2>EvaLearn: Quantifying the Learning Capability and\nEfficiency of LLMs via Sequential Problem Solving</h2>\n\n[![Paper](https://img.shields.io/badge/Paper-arXiv-blue.svg?style=for-the-badge)](https://arxiv.org/abs/2506.02672)\n[![Proceedings](https://img.shields.io/badge/Proceedings-NeurIPS%202025-8A2BE2.svg?style=for-the-badge)](https://proceedings.neurips.cc/paper_files/paper/2025/hash/b7383744ec8a316f93e235a1f7f03468-Abstract-Conference.html)\n[![Venue](https://img.shields.io/badge/Venue-NeurIPS%202025-orange.svg?style=for-the-badge)](https://neurips.cc/virtual/2025/poster/115778)\n[![Code License](https://img.shields.io/badge/Code_License-Apache_2.0-yellow.svg?style=for-the-badge)](./LICENSE)\n[![Data License](https://img.shields.io/badge/Data_License-CC_BY_4.0-red.svg?style=for-the-badge)](./DATA_LICENSE)\n\n</div>\n\n## 📰 News\n\n- **📅 Dec 4, 2025**: EvaLearn was officially published in the [NeurIPS 2025 proceedings](https://proceedings.neurips.cc/paper_files/paper/2025/hash/b7383744ec8a316f93e235a1f7f03468-Abstract-Conference.html) and presented as a poster at NeurIPS 2025 in San Diego! 🎉\n- **📅 Sep 18, 2025**: EvaLearn was accepted to the NeurIPS 2025 main track with a high score of 5/5/5/5! 🎉\n- **📅 Jul 15, 2025**: We've released a new version! 🎉 Open-sourced complete Chinese rubrics, updated Chinese README documentation, and optimized evaluation scripts for improved efficiency and accuracy.\n- **📅 Jun 5, 2025**: EvaLearn is officially open-sourced! 🚀 We released this innovative benchmark for evaluating the learning capability and efficiency of large language models.\n\n## 📚 Overview\n\nEvaLearn is a benchmark designed to evaluate large language models (LLMs) on their learning ca"},{"ref":"P19","kind":"page","title":"ByteDance-Seed/BM-code repository metadata","date":"2026-06-11T03:58:33.907058+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/BM-code","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/BM-code\n\nDescription: [Arxiv 2025] ByteMorph: Benchmarking Instruction-Guided Image Editing with Non-Rigid Motions\n\nLanguage: Python\n\nLicense: NOASSERTION\n\nStars: 45\n\nForks: 1\n\nOpen issues: 1\n\nCreated: 2025-05-09T04:57:47Z\n\nPushed: 2025-06-11T06:14:55Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<p align=\"center\">\n\n<h2 align=\"center\">ByteMorph: Benchmarking Instruction-Guided Image Editing <br> with Non-Rigid Motions</h2>\n<p align=\"center\">\n<a href=\"https://boese0601.github.io/\">Di Chang</a><sup>1,2*</sup>\n·\n<a href=\"https://github.com/ljzycmd\">Mingdeng Cao</a><sup>1,3*</sup>\n· \n<a href=\"https://seasonsh.github.io/\">Yichun Shi</a><sup>1</sup>\n· \n<a href=\"https://www.linkedin.com/in/bo-liu-340313170\">Bo Liu</a><sup>1,4</sup>\n· \n<a href=\"https://primecai.github.io/\">Shengqu Cai</a><sup>1,5</sup>\n· \n<a href=\"https://shijiezhou-ucla.github.io/\">Shijie Zhou</a><sup>6</sup>\n<br>\n<a href=\"https://scholar.google.com/citations?user=78vU1IUAAAAJ&hl=en\">Weilin Huang</a><sup>1</sup>\n· \n<a href=\"https://web.stanford.edu/~gordonwz/\">Gordon Wetzstein</a><sup>5</sup>\n· \n<a href=\"https://www.ihp-lab.org/\">Mohammad Soleymani</a><sup>2</sup>\n· \n<a href=\"https://pengwangucla.github.io/peng-wang.github.io/\">Peng Wang</a><sup>1</sup>\n<br>\n<sup>1</sup>ByteDance Seed &nbsp;<sup>2</sup>Unviersity of Southern California &nbsp; <sup>3</sup>University of Tokyo &nbsp; \n<br>\n<sup>4</sup>University of California Berkeley&nbsp; <sup>5</sup>Stanford University&nbsp; <sup>6</sup>University of California Los Angeles\n<br>\n<br>\n<sup>*</sup> denotes equal contribution\n<br>\n</br>\n<a href=\"https://arxiv.org/abs/2506.03107\">\n<img src='https://img.shields.io/badge/arXiv-Paper-red' alt='Paper PDF'></a>\n<a href='https://huggingface.co/datasets/ByteDance-Seed/BM-Bench'>\n<img src='https://img.shields.io/badge/%F0%9F%A4%97%20HuggingFace-Benchmark-yellow'></a>\n<a href='https://huggingface.co/datasets/ByteDance-Seed/BM-6M-Demo'>\n<img src='https://img.shields.io/badge/%F0%9F%A4%97%20HuggingFace-Dataset_Demo-yellow'></a>\n<a href='https://huggingface.co/datasets/ByteDance-Seed/BM-6M'>\n<img src='https://img.shields.io/badge/%F0%9F%A4%97%20HuggingFace-Dataset-yellow'></a>\n<a href='https:"},{"ref":"P20","kind":"page","title":"ByteDance-Seed/Seed1.5-VL repository metadata","date":"2026-06-11T03:58:33.787776+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Seed1.5-VL","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Seed1.5-VL\n\nDescription: Seed1.5-VL, a vision-language foundation model designed to advance general-purpose multimodal understanding and reasoning, achieving state-of-the-art performance on 38 out of 60 public benchmarks.\n\nLanguage: Jupyter Notebook\n\nLicense: Apache-2.0\n\nStars: 1580\n\nForks: 66\n\nOpen issues: 25\n\nCreated: 2025-05-11T06:00:32Z\n\nPushed: 2025-06-14T19:58:52Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n<img src=\"./assets/banner.png\" width=400>\n</div>\n\n<p align=\"center\">\n🤗 <a href=\"https://huggingface.co/spaces/ByteDance-Seed/Seed1.5-VL\"> HuggingFace Demo</a>&nbsp&nbsp | &nbsp&nbsp🌐 <a href=\"https://seed.bytedance.com/zh/tech/seed1_5_vl\"> Homepage</a>&nbsp&nbsp | &nbsp&nbsp📄 <a href=\"https://arxiv.org/abs/2505.07062\">arXiv</a>\n</p>\n\nToday, we are excited to introduce **Seed1.5-VL** 🚀, a powerful and efficient vision-language foundation model designed for advanced general-purpose multimodal understanding and reasoning.\n\n## 🌟 Highlights\n* 🧠 **Efficient Powerhouse:** Achieves top performance with a relatively modest architecture, 532M vision encoder & 20B active parameter MoE LLM.\n* 🏆 **Exceptional Benchmark Performance:** Delivers State-of-the-Art results on 38 out of 60 public VLM benchmarks, demonstrating broad competence.\n* 💡 **Versatile Capabilities:** Excels across diverse capabilities including complex reasoning (e.g., visual puzzles like Rebus), OCR, diagram understanding, visual grounding, 3D spatial understanding, and video comprehension.\n* 🤖 **Advanced Agent-Centric Abilities:** Demonstrates leading performance in interactive agent tasks, showcasing strong capabilities in GUI control and gameplay.\n\n**This repository offers usage cookbook and best practices designed to help developers effectively use Seed1.5-VL.**\n\n## 📢 News\n* `2025-05-13:` We have deployed our Seed1.5-VL on [🤗 HuggingFace Spaces](https://huggingface.co/spaces/ByteDance-Seed/Seed1.5-VL), Welcome to try out our model!\n* `2025-05-12:` We have released the [Seed1.5-VL Technical Report](./Seed1.5-VL-Technical-Report.pdf).\n* `2025-05-12:` We are extremely delighted to release the flagship Seed1.5-VL on [Volcano Engine](https"},{"ref":"P21","kind":"page","title":"ByteDance-Seed/SeedVR repository metadata","date":"2026-06-11T03:58:33.632869+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/SeedVR","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/SeedVR\n\nDescription: Repo for SeedVR2 (ICLR2026) & SeedVR (CVPR2025 Highlight)\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 1222\n\nForks: 71\n\nOpen issues: 46\n\nCreated: 2025-06-10T10:41:54Z\n\nPushed: 2026-01-27T06:02:24Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n<img src=\"assets/seedvr_logo.png\" alt=\"SeedVR\" width=\"400\"/>\n</div>\n\n# SeedVR: Seeding Infinity in Diffusion Transformer Towards Generic Video Restoration\n> [Jianyi Wang](https://iceclear.github.io), [Zhijie Lin](https://scholar.google.com/citations?user=xXMj6_EAAAAJ&hl=zh-CN), [Meng Wei](https://openreview.net/profile?id=~Meng_Wei11), [Ceyuan Yang](https://scholar.google.com/citations?user=uPmTOHAAAAAJ&hl=zh-CN), [Fei Xiao](https://openreview.net/profile?id=~Fei_xiao8), [Chen Change Loy](https://www.mmlab-ntu.com/person/ccloy/), [Lu Jiang](http://www.lujiang.info/)\n>\n> **CVPR 2025 (Highlight)**\n\n<p>\n<a href=\"https://iceclear.github.io/projects/seedvr/\">\n<img\nsrc=\"https://img.shields.io/badge/SeedVR-Website-0A66C2?logo=safari&logoColor=white\"\nalt=\"SeedVR Website\"\n/>\n</a>\n<a href=\"https://huggingface.co/collections/ByteDance-Seed/seedvr-6849deeb461c4e425f3e6f9e\">\n<img \nsrc=\"https://img.shields.io/badge/SeedVR-Models-yellow?logo=huggingface&logoColor=yellow\" \nalt=\"SeedVR Models\"\n/>\n</a>\n<a href=\"https://huggingface.co/spaces/ByteDance-Seed/SeedVR2-3B\">\n<img \nsrc=\"https://img.shields.io/badge/SeedVR2-Space-orange?logo=huggingface&logoColor=yellow\" \nalt=\"SeedVR2 Space\"\n/>\n</a>\n<a href=\"https://github.com/numz/ComfyUI-SeedVR2_VideoUpscaler\">\n<img\nsrc=\"https://img.shields.io/badge/SeedVR-ComfyUI-blue?logo=googleplay&logoColor=blue\"\nalt=\"SeedVR ComfyUI\"\n/>\n</a>\n<a href=\"https://arxiv.org/abs/2501.01320\">\n<img\nsrc=\"https://img.shields.io/badge/SeedVR-Paper-red?logo=arxiv&logoColor=red\"\nalt=\"SeedVR Paper on ArXiv\"\n/>\n</a>\n<a href=\"https://www.youtube.com/watch?v=aPpBs_B2iCY\" target='_blank'>\n<img \nsrc=\"https://img.shields.io/badge/Demo%20Video-%23FF0000.svg?logo=YouTube&logoColor=white\"\nalt=\"SeedVR Video Demo on YouTube\"\n/>\n</a>\n</p>\n\n>\n> **Why SeedVR:** Conventional restoration models achieve inferior performance on both real-world and AIGC video restoration due "},{"ref":"P22","kind":"page","title":"ByteDance-Seed/VINCIE repository metadata","date":"2026-06-11T03:58:33.119362+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/VINCIE","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/VINCIE\n\nDescription: Official code for VINCIE: Unlocking In-context Image Editing from Video\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 57\n\nForks: 4\n\nOpen issues: 4\n\nCreated: 2025-06-30T03:50:56Z\n\nPushed: 2026-03-28T09:43:48Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# VINCIE: Unlocking In-context Image Editing from Video\n<p align=\"center\">\n<a href=\"https://vincie2025.github.io/\">\n<img\nsrc=\"https://img.shields.io/badge/VINCIE-Website-0A66C2?logo=safari&logoColor=white\"\nalt=\"VINCIE Website\"\n/>\n</a>\n<a href=\"https://arxiv.org/abs/2506.10941\">\n<img\nsrc=\"https://img.shields.io/badge/VINCIE-Paper-red?logo=arxiv&logoColor=red\"\nalt=\"VINCIE Paper on ArXiv\"\n/>\n</a>\n<a href=\"https://github.com/ByteDance-Seed/VINCIE\">\n<img \nalt=\"Github\" src=\"https://img.shields.io/badge/VINCIE-Codebase-536af5?color=536af5&logo=github\"\nalt=\"VINCIE Codebase\"\n/>\n</a>\n<a href=\"https://huggingface.co/collections/ByteDance-Seed/vincie-6864cc2e3116d82e4a83a17c\">\n<img \nsrc=\"https://img.shields.io/badge/VINCIE-Models-yellow?logo=huggingface&logoColor=yellow\" \nalt=\"VINCIE Models\"\n/>\n</a>\n<a href=\"https://huggingface.co/datasets/leigangqu/VINCIE-10M\">\n<img \nsrc=\"https://img.shields.io/badge/VINCIE (10M)-Dataset-yellow?logo=huggingface&logoColor=yellow\" \nalt=\"VINCIE-10M Dataset\"\n/>\n</a>\n<a href=\"https://huggingface.co/datasets/leigangqu/MSE-Bench\">\n<img \nsrc=\"https://img.shields.io/badge/MSE-Benchmark-yellow?logo=huggingface&logoColor=yellow\" \nalt=\"MSE-Bench\"\n/>\n</a>\n<!-- <a href=\"https://huggingface.co/spaces/ByteDance-Seed/VINCIE-3B\">\n<img \nsrc=\"https://img.shields.io/badge/VINCIE-Space-orange?logo=huggingface&logoColor=yellow\" \nalt=\"VINCIE Space\"\n/>\n</a> -->\n</p>\n\n> [Leigang Qu](https://leigang-qu.github.io/), [Feng Cheng](https://klauscc.github.io/), [Ziyan Yang](https://ziyanyang.github.io/), [Qi Zhao](https://kevinz8866.github.io/), [Shanchuan Lin](https://scholar.google.com/citations?user=EDWUw7gAAAAJ&hl=en), [Yichun Shi](https://seasonsh.github.io/), [Yicong Li](https://yl3800.github.io/), [Wenjie Wang](https://wenjiewwj.github.io/), [Tat-Seng Chua](https://www.chuatatseng.com/), [Lu Jiang](http://www.lujiang.info/index.html)\n> \n> In-context image editing aims"},{"ref":"P23","kind":"page","title":"ByteDance-Seed/Seed-X-7B repository metadata","date":"2026-06-11T03:58:33.049129+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Seed-X-7B","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Seed-X-7B\n\nLanguage: Python\n\nLicense: NOASSERTION\n\nStars: 172\n\nForks: 7\n\nOpen issues: 15\n\nCreated: 2025-07-16T07:18:16Z\n\nPushed: 2025-08-18T11:40:20Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone! \n<br>\nWe are <b>ByteDance Seed team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://seed.bytedance.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n<a href=\"https://github.com/user-attachments/assets/5793e67c-79bb-4a59-811a-fcc7ed510bd4\">\n<img src=\"https://img.shields.io/badge/WeChat-07C160?style=for-the-badge&logo=wechat&logoColor=white\"></a>\n<a href=\"https://www.xiaohongshu.com/user/profile/668e7e15000000000303157d?xsec_token=ABl2-aqekpytY6A8TuxjrwnZskU-6BsMRE_ufQQaSAvjc%3D&xsec_source=pc_search\">\n<img src=\"https://img.shields.io/badge/Xiaohongshu-%23FF2442?style=for-the-badge&logo=xiaohongshu&logoColor=white\"></a>\n<a href=\"https://www.zhihu.com/org/dou-bao-da-mo-xing-tuan-dui/\">\n<img src=\"https://img.shields.io/badge/zhihu-%230084FF?style=for-the-badge&logo=zhihu&logoColor=white\"></a>\n</p>\n\n<div align=center>\n<img src=\"https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216\"/></div>\n</div>\n\n<!-- 注释：以上为Seed官方信息，可直接复制使用，请注意导入“Seed WeChat”（第12行）、“Seed logo”(第20行)图片替换 -->\n\n# Seed-X: Building Strong Multilingual Translation LLM with 7B Parameters\n<p align=\"center\">\n<a href=\"https://arxiv.org/pdf/2507.13618\">\n<img src=\"https://img.shields.io/badge/Seed--X-Report-blue\"></a>\n<a href=\"https://huggingface.co/collections/ByteDance-Seed/seed-x-6878753f2858bc17afa78543\">\n<img src=\"https://img.shields.io/badge/Seed--X-Hugging Face-brightgreen\"></a>\n<a href=\"https://huggingface.co/spaces/ByteDance-Seed/Seed-X\">\n<img src=\"https://img.shields.io/badge/Seed--X-DEMO-purple\"></a>\n<a href=\"https://github.com/ByteDance-Seed/Seed-X-7B/blob/main/LICENSE.openmdw\">\n<img src=\"https://img.shields.io/badge/License-OpenMDW-yellow\"></a>\n</p>\n\n<!-- 🤗 [HuggingFace]() | 📄 [Technical Report](/Technical_Report.pdf) -->\n\nWe are excited to introduce **Seed-X**, a powerful serie"},{"ref":"P24","kind":"page","title":"ByteDance-Seed/Seed-Prover repository metadata","date":"2026-06-11T03:58:32.935473+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/Seed-Prover","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/Seed-Prover\n\nLanguage: Lean\n\nLicense: Apache-2.0\n\nStars: 433\n\nForks: 28\n\nOpen issues: 3\n\nCreated: 2025-07-16T07:37:21Z\n\nPushed: 2026-02-13T15:03:35Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone! \n<br>\nWe are <b>ByteDance Seed team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://team.doubao.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n<a href=\"https://github.com/user-attachments/assets/93481cda-a7f3-47f3-b333-fe6b3da86b78\">\n<img src=\"https://img.shields.io/badge/WeChat-07C160?style=for-the-badge&logo=wechat&logoColor=white\"></a>\n<a href=\"Seed研究员 - 小红书\">\n<img src=\"https://img.shields.io/badge/Xiaohongshu-%23FF2442?style=for-the-badge&logo=xiaohongshu&logoColor=white\"></a>\n<a href=\"https://www.zhihu.com/org/dou-bao-da-mo-xing-tuan-dui/\">\n<img src=\"https://img.shields.io/badge/zhihu-%230084FF?style=for-the-badge&logo=zhihu&logoColor=white\"></a>\n</p>\n\n![seed logo](https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216)\n\n# <img src=\"./imgs/logo.png\" height=\"25\"> Seed-Prover 1.5\n\nThis page is used to store the Seed AI4Math group’s research projects, including Seed‑Prover 1.5, Seed‑Prover and Delta‑Prover.\n- **Seed Prover 1.5** Seed-Prover 1.5 is our latest formal proving system. It can solve 88% of the problems in PutnamBench and 11 out of the 12 competition problems from Putnam 2025. [Arxiv](https://arxiv.org/abs/2512.17260)\n- **Seed Prover** Seed‑Prover 1.0 is the system we officially participated with in the IMO 2025. [Arxiv](https://arxiv.org/abs/2507.23726)\n- **Delta prover** Delta‑Prover is a separate project focused on researching test time techniques for generating formal proofs. [Arxiv](https://arxiv.org/abs/2507.15225)\n\n## Seed Prover IMO 2025\nSeed Prover 1.0 solved 4 out of 6 problems in IMO 2025 durint the context, with the following breakdown:\n- **Day 1**: Fully solved P2 (geometry) and P3 (number theory), fully solved P1 (combinatorics) after the competition.\n- **Day 2**: Fully solved P4 (number theory) and P5 (combinatorics "},{"ref":"P25","kind":"page","title":"ByteDance-Seed/manip-as-in-sim-suite repository metadata","date":"2026-06-11T03:58:32.878666+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/manip-as-in-sim-suite","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/manip-as-in-sim-suite\n\nDescription: Sim-to-real and CDM inference code for ManipAsInSim project.\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 154\n\nForks: 6\n\nOpen issues: 9\n\nCreated: 2025-08-08T09:39:36Z\n\nPushed: 2025-12-09T08:31:31Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# Manipulation As in Simulation Suite\n\nThis repository contains the open-source implementation of the paper **\"Manipulation as in Simulation: Enabling Accurate Geometry Perception in Robots\"**. The suite provides tools for bridging the sim-to-real gap in robotic manipulation through high-quality depth perception and automated demonstration generation.\n\n## 📋 Table of Contents\n\n- [🎯 Overview](#-overview)\n- [🔍 CDM (Camera Depth Models)](#-cdm-camera-depth-models)\n- [🤖 WBCMimic](#-wbcmimic)\n- [✨ Key Features](#-key-features)\n- [🚀 Quick Start](#-quick-start)\n- [Prerequisites](#prerequisites)\n- [Installation](#installation)\n- [🔧 Environment Setup](#-environment-setup)\n- [Environment Variables](#environment-variables)\n- [Setup Options](#setup-options)\n- [Verification](#verification)\n- [Troubleshooting](#troubleshooting)\n- [🚀 Usage](#-usage)\n- [CDM Usage](#cdm-usage)\n- [WBCMimic Usage](#wbcmimic-usage)\n- [🔬 Research Contributions](#-research-contributions)\n- [Camera Depth Models (CDMs)](#camera-depth-models-cdms)\n- [WBCMimic Enhancements](#wbcmimic-enhancements)\n- [🎯 Supported Tasks & Hardware](#-supported-tasks--hardware)\n- [Robotic Tasks](#robotic-tasks)\n- [Supported Cameras](#supported-cameras)\n- [📚 Documentation](#-documentation)\n- [📄 Citation](#-citation)\n- [📝 License](#-license)\n- [🔗 Links](#-links)\n- [📧 Contact](#-contact)\n\n## 🎯 Overview\n\nThe suite consists of two main components that enable seamless sim-to-real transfer for robotic manipulation:\n\n### 🔍 CDM (Camera Depth Models)\nA depth estimation package that produces clean, simulation-like depth maps from noisy real-world camera data. CDMs enable policies trained purely in simulation to transfer directly to real robots by providing perfect depth perception.\n\n### 🤖 WBCMimic \nAn enhanced version of MimicGen that extends autonomous data generation to mobile manipulators with whole-body control. I"},{"ref":"P26","kind":"page","title":"ByteDance-Seed/cudaLLM repository metadata","date":"2026-06-11T03:58:32.376919+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/cudaLLM","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/cudaLLM\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 144\n\nForks: 7\n\nOpen issues: 3\n\nCreated: 2025-08-01T22:30:11Z\n\nPushed: 2025-08-18T06:56:11Z\n\nDefault branch: main\n\nFork: no\n\nArchived: no\n\nREADME:\n# CudaLLM: Training Language Models to Generate High-Performance CUDA Kernels\n\n<a href=\"https://huggingface.co/ByteDance-Seed/cudaLLM-8B\">\n<img \nsrc=\"https://img.shields.io/badge/CudaLLM-Model-yellow?logo=huggingface&logoColor=yellow\" \nalt=\"CudaLLM Model\"\n/>\n</a>\n\n<a href=\"https://huggingface.co/datasets/ByteDance-Seed/cudaLLM-data\">\n<img \nsrc=\"https://img.shields.io/badge/CudaLLM-Data-blue?logo=huggingface&logoColor=yellow\" \nalt=\"CudaLLM Data\"\n/>\n</a>\n\nThis project provides a complete pipeline for training LLMs to automatically generate efficient and correct CUDA kernels. By leveraging a two-stage process of SFT and RL, this framework fine-tunes a base model to write optimized CUDA code.\n\nFor demonstration purposes, this guide uses [Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B) as the base model.\n\n## How It Works\n\nThe training methodology is composed of two main stages:\n\n1. **SFT:** The base LLM is first fine-tuned on a high-quality dataset of CUDA kernel examples. The data is generated by DeepSeek R1, DeepSeel Coder-7B, and Qwen2-32B.\n2. **RL:** After SFT, the model is further optimized through reinforcement learning. In this stage, the model generates CUDA kernels which are then compiled and tested. This feedback signal is used as a reward to train the model to produce valid kernels.\n\n## Getting Started\nTo set up and run the training pipeline, follow these steps:\n### Step 0: Prepare Datasets\n\nFirst, you need to process the raw datasets for SFT and RL, and download the evaluation dataset. This script handles the necessary preprocessing.\n\n* **SFT Dataset:** `sft_cuda_llm_r1.parquet`\n* **RL Dataset:** `rl_cuda_llm_0424.parquet`\n* **Evaluation Dataset:** [KernelBench](https://huggingface.co/datasets/ScalingIntelligence/KernelBench)\n\nRun the following command to begin:\n\n```bash\n# install verl, the git SHA is abb87bc147467589d1357dd80a1e3fefa188e11f\ngit clone https://github.com/volcengine/verl.git\ncd verl\npip install --no-deps -e .\ncd ..\n\npython3 cu"},{"ref":"P27","kind":"page","title":"ByteDance-Seed/seed-oss repository metadata","date":"2026-06-11T03:58:32.349866+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/seed-oss","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/seed-oss\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 885\n\nForks: 50\n\nOpen issues: 10\n\nCreated: 2025-08-08T00:48:59Z\n\nPushed: 2025-09-15T03:44:13Z\n\nDefault branch: master\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=\"center\">\n👋 Hi, everyone!\n<br>\nWe are <b>ByteDance Seed Team.</b>\n</div>\n\n<p align=\"center\">\nYou can get to know us better through the following channels👇\n<br>\n<a href=\"https://seed.bytedance.com/\">\n<img src=\"https://img.shields.io/badge/Website-%231e37ff?style=for-the-badge&logo=bytedance&logoColor=white\"></a>\n</p>\n\n![seed logo](https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216)\n\n# Seed-OSS Open-Source Models\n<p align=\"center\">\n<a href=\"https://github.com/ByteDance-Seed/seed-oss\">\n<img src=\"https://img.shields.io/badge/Seed-Project Page-yellow\"></a>\n<a href=\"https://github.com/ByteDance-Seed/seed-oss\">\n<img src=\"https://img.shields.io/badge/Seed-Tech Report Coming Soon-red\"></a>\n<a href=\"https://huggingface.co/collections/ByteDance-Seed/seed-oss-68a609f4201e788db05b5dcd\">\n<img src=\"https://img.shields.io/badge/Seed-Hugging Face-orange\"></a>\n<br>\n<a href=\"./LICENSE\">\n<img src=\"https://img.shields.io/badge/License-Apache2.0-blue\"></a>\n</p>\n\n## News\n- [2025/08/20]🔥We release `Seed-OSS-36B-Base` (both with and without synthetic data versions) and `Seed-OSS-36B-Instruct`.\n\n## Introduction\nSeed-OSS is a series of open-source large language models developed by ByteDance's Seed Team, designed for powerful long-context, reasoning, agent and general capabilities, and versatile developer-friendly features. Although trained with only 12T tokens, Seed-OSS achieves excellent performance on several popular open benchmarks.\n\nWe release this series of models to the open-source community under the Apache-2.0 license.\n\n> [!NOTE]\n> Seed-OSS is primarily optimized for international (i18n) use cases.\n\n### Key Features\n- **Flexible Control of Thinking Budget**: Allowing users to flexibly adjust the reasoning length as needed. This capability of dynamically controlling the reasoning length enhances inference efficiency in practical application scenarios.\n- **Enhanced Reasoning Capability**: Specifically optimized for reaso"},{"ref":"P28","kind":"page","title":"ByteDance-Seed/m3-agent repository metadata","date":"2026-06-11T03:58:32.303449+00:00","date_source":null,"source_url":"https://github.com/ByteDance-Seed/m3-agent","signal_url":null,"signal_json_url":null,"text":"# ByteDance-Seed/m3-agent\n\nLanguage: Python\n\nLicense: Apache-2.0\n\nStars: 1378\n\nForks: 113\n\nOpen issues: 17\n\nCreated: 2025-07-30T13:12:32Z\n\nPushed: 2026-02-12T06:03:56Z\n\nDefault branch: master\n\nFork: no\n\nArchived: no\n\nREADME:\n<div align=left>\n<img src=\"https://github.com/user-attachments/assets/c42e675e-497c-4508-8bb9-093ad4d1f216\" width=40%>\n</div>\n\n<h1 style=\"text-align: center;\">Seeing, Listening, Remembering, and Reasoning: A Multimodal Agent with Long-Term Memory<br>ICLR 2026</h1>\n\n[![arXiv](https://img.shields.io/badge/arXiv-2508.09736-b31b1b.svg)](https://arxiv.org/abs/2508.09736)\n[![Demo](https://img.shields.io/badge/homepage-M3--Agent-blue)](https://m3-agent.github.io)\n[![Model](https://img.shields.io/badge/model_HF-Memorization-green)](https://huggingface.co/ByteDance-Seed/M3-Agent-Memorization)\n[![Model](https://img.shields.io/badge/model_HF-Control-darkgreen)](https://huggingface.co/ByteDance-Seed/M3-Agent-Control)\n[![Data](https://img.shields.io/badge/data-M3--Bench-F9D371)](https://huggingface.co/datasets/ByteDance-Seed/M3-Bench)\n\n## Abstract\n\nWe introduce M3-Agent, a novel multimodal agent framework equipped with long-term memory. Like humans, M3-Agent can process real-time visual and auditory inputs to build and update its long-term memory. Beyond episodic memory, it also develops semantic memory, enabling it to accumulate world knowledge over time. Its memory is organized in an entity-centric, multimodal format, allowing deeper and more consistent understanding of the environment. Given an instruction, M3-Agent autonomously performs multi-turn, iterative reasoning and retrieves relevant information from memory to accomplish the task. To evaluate memory effectiveness and memory-based reasoning in multimodal agents, we develop M3-Bench, a new long-video question answering benchmark. M3-Bench comprises 100 newly recorded real-world videos captured from a robot’s perspective (M3-Bench-robot) and 920 web-sourced videos across diverse scenarios (M3-Bench-web). We annotate question-answer pairs designed to test key capabilities essential for agent applications, such as human understanding, general knowledge extraction, and cross- modal reasoning. Experi"},{"ref":"E1","kind":"event","title":"ByteDance-Seed/Seed-OSS-36B-Instruct","date":"2025-08-20T15:03:26+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct","signal_url":"https://onlylabs.fyi/signals/6f3af2c1-697e-42f4-989b-2f3a707dc811","signal_json_url":"https://onlylabs.fyi/signals/6f3af2c1-697e-42f4-989b-2f3a707dc811/signal.json","text":"model_released · ByteDance-Seed/Seed-OSS-36B-Instruct · signal_desk=releases · occurred_at=2025-08-20T15:03:26+00:00 · url=https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct · hf_downloads=36952 · hf_likes=500 · hf_params=36151104512 · pipeline=text-generation · license=apache-2.0"},{"ref":"E2","kind":"event","title":"ByteDance-Seed/Stable-DiffCoder-8B-Instruct","date":"2026-01-15T07:49:01+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/Stable-DiffCoder-8B-Instruct","signal_url":"https://onlylabs.fyi/signals/4c28f3ad-c7fa-4484-b740-27b5e098fde1","signal_json_url":"https://onlylabs.fyi/signals/4c28f3ad-c7fa-4484-b740-27b5e098fde1/signal.json","text":"model_released · ByteDance-Seed/Stable-DiffCoder-8B-Instruct · signal_desk=releases · occurred_at=2026-01-15T07:49:01+00:00 · url=https://huggingface.co/ByteDance-Seed/Stable-DiffCoder-8B-Instruct · hf_downloads=927 · hf_likes=137 · hf_params=8250462208 · pipeline=text-generation · license=mit · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E3","kind":"event","title":"ByteDance-Seed/Cola-DLM","date":"2026-05-15T07:55:26+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/Cola-DLM","signal_url":"https://onlylabs.fyi/signals/a6b59621-a4bd-4f69-a1c7-81dd4c45f526","signal_json_url":"https://onlylabs.fyi/signals/a6b59621-a4bd-4f69-a1c7-81dd4c45f526/signal.json","text":"model_released · ByteDance-Seed/Cola-DLM · signal_desk=releases · occurred_at=2026-05-15T07:55:26+00:00 · url=https://huggingface.co/ByteDance-Seed/Cola-DLM · hf_likes=36 · pipeline=text-generation · license=apache-2.0"},{"ref":"E4","kind":"event","title":"ByteDance-Seed/Seed-OSS-36B-Base","date":"2025-08-20T13:57:44+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Base","signal_url":"https://onlylabs.fyi/signals/96f63b61-e5ee-4703-ac28-ee966ce64cfe","signal_json_url":"https://onlylabs.fyi/signals/96f63b61-e5ee-4703-ac28-ee966ce64cfe/signal.json","text":"model_released · ByteDance-Seed/Seed-OSS-36B-Base · signal_desk=releases · occurred_at=2025-08-20T13:57:44+00:00 · url=https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Base · hf_downloads=1692 · hf_likes=62 · hf_params=36151104512 · pipeline=text-generation · license=apache-2.0"},{"ref":"E5","kind":"event","title":"ByteDance-Seed/Seed-OSS-36B-Base-woSyn","date":"2025-08-20T14:49:52+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Base-woSyn","signal_url":"https://onlylabs.fyi/signals/4d81abea-283f-4a47-9f3d-2efec88dd215","signal_json_url":"https://onlylabs.fyi/signals/4d81abea-283f-4a47-9f3d-2efec88dd215/signal.json","text":"model_released · ByteDance-Seed/Seed-OSS-36B-Base-woSyn · signal_desk=releases · occurred_at=2025-08-20T14:49:52+00:00 · url=https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Base-woSyn · hf_downloads=114 · hf_likes=53 · hf_params=36151104512 · pipeline=text-generation · license=apache-2.0"},{"ref":"E6","kind":"event","title":"ByteDance-Seed/M3-Agent-Control","date":"2025-08-12T11:05:07+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/M3-Agent-Control","signal_url":"https://onlylabs.fyi/signals/1d3da746-d4df-44fc-afb6-068e257ba384","signal_json_url":"https://onlylabs.fyi/signals/1d3da746-d4df-44fc-afb6-068e257ba384/signal.json","text":"model_released · ByteDance-Seed/M3-Agent-Control · signal_desk=releases · occurred_at=2025-08-12T11:05:07+00:00 · url=https://huggingface.co/ByteDance-Seed/M3-Agent-Control · hf_downloads=51 · hf_likes=50 · hf_params=32762123264 · license=apache-2.0"},{"ref":"E7","kind":"event","title":"ByteDance-Seed/TaskMem","date":"2026-06-02T03:10:24+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/TaskMem","signal_url":"https://onlylabs.fyi/signals/531e8c80-7cb1-412c-a757-f02e544b66bd","signal_json_url":"https://onlylabs.fyi/signals/531e8c80-7cb1-412c-a757-f02e544b66bd/signal.json","text":"model_released · ByteDance-Seed/TaskMem · signal_desk=releases · occurred_at=2026-06-02T03:10:24+00:00 · url=https://huggingface.co/ByteDance-Seed/TaskMem · hf_downloads=29 · hf_likes=4 · hf_params=31070754032 · license=apache-2.0"},{"ref":"E8","kind":"event","title":"ByteDance-Seed/TaskMem","date":"2026-05-29T03:03:17+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/TaskMem","signal_url":"https://onlylabs.fyi/signals/55e0369a-bd79-48f1-a004-0cd4fdfb03e4","signal_json_url":"https://onlylabs.fyi/signals/55e0369a-bd79-48f1-a004-0cd4fdfb03e4/signal.json","text":"repo_new · ByteDance-Seed/TaskMem · signal_desk=repos · occurred_at=2026-05-29T03:03:17+00:00 · url=https://github.com/ByteDance-Seed/TaskMem · stars=17 · raw={\"repo\":\"ByteDance-Seed/TaskMem\",\"language\":\"Python\"}"},{"ref":"E9","kind":"event","title":"ByteDance-Seed/SimArt","date":"2026-05-19T08:55:39+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/SimArt","signal_url":"https://onlylabs.fyi/signals/4add7ea8-8980-4072-842f-2f2bc5bf9a84","signal_json_url":"https://onlylabs.fyi/signals/4add7ea8-8980-4072-842f-2f2bc5bf9a84/signal.json","text":"model_released · ByteDance-Seed/SimArt · signal_desk=releases · occurred_at=2026-05-19T08:55:39+00:00 · url=https://huggingface.co/ByteDance-Seed/SimArt · hf_likes=4 · license=apache-2.0"},{"ref":"E10","kind":"event","title":"ByteDance-Seed/VeOmni v0.1.11","date":"2026-05-26T07:42:38+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.11","signal_url":"https://onlylabs.fyi/signals/a295b741-e912-4ff5-a12a-263f6926fbd4","signal_json_url":"https://onlylabs.fyi/signals/a295b741-e912-4ff5-a12a-263f6926fbd4/signal.json","text":"release · ByteDance-Seed/VeOmni v0.1.11 · signal_desk=releases · occurred_at=2026-05-26T07:42:38+00:00 · url=https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.11 · raw={\"repo\":\"ByteDance-Seed/VeOmni\"}"},{"ref":"E11","kind":"event","title":"ByteDance-Seed/VeOmni v0.1.10","date":"2026-05-21T07:21:47+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.10","signal_url":"https://onlylabs.fyi/signals/b0dfde85-a08b-4b5d-b63f-3f8075c3c0a5","signal_json_url":"https://onlylabs.fyi/signals/b0dfde85-a08b-4b5d-b63f-3f8075c3c0a5/signal.json","text":"release · ByteDance-Seed/VeOmni v0.1.10 · signal_desk=releases · occurred_at=2026-05-21T07:21:47+00:00 · url=https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.10 · raw={\"repo\":\"ByteDance-Seed/VeOmni\"}"},{"ref":"E12","kind":"event","title":"ByteDance-Seed/Cola-DLM","date":"2026-05-15T07:09:04+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Cola-DLM","signal_url":"https://onlylabs.fyi/signals/957704c5-24c9-4360-aafa-f8382010ed1d","signal_json_url":"https://onlylabs.fyi/signals/957704c5-24c9-4360-aafa-f8382010ed1d/signal.json","text":"repo_new · ByteDance-Seed/Cola-DLM · signal_desk=repos · occurred_at=2026-05-15T07:09:04+00:00 · url=https://github.com/ByteDance-Seed/Cola-DLM · stars=219 · raw={\"repo\":\"ByteDance-Seed/Cola-DLM\",\"description\":\"The codebase of Cola DLM\",\"language\":\"Python\"}"},{"ref":"E13","kind":"event","title":"ByteDance-Seed/THEMol","date":"2026-05-15T05:36:08+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/THEMol","signal_url":"https://onlylabs.fyi/signals/22ca2261-1970-4664-95e2-1c4d404c0eb3","signal_json_url":"https://onlylabs.fyi/signals/22ca2261-1970-4664-95e2-1c4d404c0eb3/signal.json","text":"repo_new · ByteDance-Seed/THEMol · signal_desk=repos · occurred_at=2026-05-15T05:36:08+00:00 · url=https://github.com/ByteDance-Seed/THEMol · stars=12 · raw={\"repo\":\"ByteDance-Seed/THEMol\",\"language\":\"Python\"}"},{"ref":"E14","kind":"event","title":"ByteDance-Seed/Bagel","date":"2025-04-17T06:54:07+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Bagel","signal_url":"https://onlylabs.fyi/signals/6b89dd92-0616-4428-8ff8-e9ce649e96c9","signal_json_url":"https://onlylabs.fyi/signals/6b89dd92-0616-4428-8ff8-e9ce649e96c9/signal.json","text":"repo_new · ByteDance-Seed/Bagel · signal_desk=repos · occurred_at=2025-04-17T06:54:07+00:00 · url=https://github.com/ByteDance-Seed/Bagel · stars=6000 · raw={\"repo\":\"ByteDance-Seed/Bagel\",\"description\":\"Open-source unified multimodal model\",\"language\":\"Python\"}"},{"ref":"E15","kind":"event","title":"ByteDance-Seed/cudaLLM-8B","date":"2025-08-03T00:40:02+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/cudaLLM-8B","signal_url":"https://onlylabs.fyi/signals/52038a39-ade8-4e6a-90b1-cfbaf97f3ae1","signal_json_url":"https://onlylabs.fyi/signals/52038a39-ade8-4e6a-90b1-cfbaf97f3ae1/signal.json","text":"model_released · ByteDance-Seed/cudaLLM-8B · signal_desk=releases · occurred_at=2025-08-03T00:40:02+00:00 · url=https://huggingface.co/ByteDance-Seed/cudaLLM-8B · hf_downloads=174 · hf_likes=29 · hf_params=8190735360 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E16","kind":"event","title":"ByteDance-Seed/Depth-Anything-3","date":"2025-11-12T08:44:03+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Depth-Anything-3","signal_url":"https://onlylabs.fyi/signals/e3745378-0c95-43a3-8cea-cb983a3cfb28","signal_json_url":"https://onlylabs.fyi/signals/e3745378-0c95-43a3-8cea-cb983a3cfb28/signal.json","text":"repo_new · ByteDance-Seed/Depth-Anything-3 · signal_desk=repos · occurred_at=2025-11-12T08:44:03+00:00 · url=https://github.com/ByteDance-Seed/Depth-Anything-3 · stars=5534 · raw={\"repo\":\"ByteDance-Seed/Depth-Anything-3\",\"description\":\"Depth Anything 3\",\"language\":\"Python\"}"},{"ref":"E17","kind":"event","title":"ByteDance-Seed/VeOmni v0.1.9a5","date":"2026-05-07T12:42:41+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.9a5","signal_url":"https://onlylabs.fyi/signals/e594588f-cf2f-434b-b86f-cdc529fef4ce","signal_json_url":"https://onlylabs.fyi/signals/e594588f-cf2f-434b-b86f-cdc529fef4ce/signal.json","text":"release · ByteDance-Seed/VeOmni v0.1.9a5 · signal_desk=releases · occurred_at=2026-05-07T12:42:41+00:00 · url=https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.9a5 · raw={\"repo\":\"ByteDance-Seed/VeOmni\"}"},{"ref":"E18","kind":"event","title":"ByteDance-Seed/felis","date":"2026-05-06T08:54:43+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/felis","signal_url":"https://onlylabs.fyi/signals/6a0854bb-9654-4e4e-92dc-e8fff1d6c7b9","signal_json_url":"https://onlylabs.fyi/signals/6a0854bb-9654-4e4e-92dc-e8fff1d6c7b9/signal.json","text":"repo_new · ByteDance-Seed/felis · signal_desk=repos · occurred_at=2026-05-06T08:54:43+00:00 · url=https://github.com/ByteDance-Seed/felis · stars=5 · raw={\"repo\":\"ByteDance-Seed/felis\",\"description\":\"FELIS: Free Energy of Ligand-protein InteractionS\",\"language\":\"Python\"}"},{"ref":"E19","kind":"event","title":"ByteDance-Seed/VeOmni v0.1.9a4","date":"2026-05-05T18:21:23+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.9a4","signal_url":"https://onlylabs.fyi/signals/af7b1780-571b-4858-8115-82205e4f3aed","signal_json_url":"https://onlylabs.fyi/signals/af7b1780-571b-4858-8115-82205e4f3aed/signal.json","text":"release · ByteDance-Seed/VeOmni v0.1.9a4 · signal_desk=releases · occurred_at=2026-05-05T18:21:23+00:00 · url=https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.9a4 · raw={\"repo\":\"ByteDance-Seed/VeOmni\"}"},{"ref":"E20","kind":"event","title":"ByteDance-Seed/VeOmni v0.1.9a3","date":"2026-05-01T21:00:32+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.9a3","signal_url":"https://onlylabs.fyi/signals/df509342-e36d-40aa-a3f2-eda2b457b28e","signal_json_url":"https://onlylabs.fyi/signals/df509342-e36d-40aa-a3f2-eda2b457b28e/signal.json","text":"release · ByteDance-Seed/VeOmni v0.1.9a3 · signal_desk=releases · occurred_at=2026-05-01T21:00:32+00:00 · url=https://github.com/ByteDance-Seed/VeOmni/releases/tag/v0.1.9a3 · raw={\"repo\":\"ByteDance-Seed/VeOmni\"}"},{"ref":"E21","kind":"event","title":"ByteDance-Seed/SimArt","date":"2026-04-23T03:25:52+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/SimArt","signal_url":"https://onlylabs.fyi/signals/05e5f295-f84c-41c4-89df-baa3606008b1","signal_json_url":"https://onlylabs.fyi/signals/05e5f295-f84c-41c4-89df-baa3606008b1/signal.json","text":"repo_new · ByteDance-Seed/SimArt · signal_desk=repos · occurred_at=2026-04-23T03:25:52+00:00 · url=https://github.com/ByteDance-Seed/SimArt · stars=3 · raw={\"repo\":\"ByteDance-Seed/SimArt\",\"description\":\"[SIGGRAPH 2026] SIMART: Decomposing Monolithic Meshes into Sim-ready Articulated Assets via MLLM\",\"language\":\"Python\"}"},{"ref":"E22","kind":"event","title":"ByteDance-Seed/Stable-DiffCoder-8B-Base","date":"2026-01-15T07:48:10+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/Stable-DiffCoder-8B-Base","signal_url":"https://onlylabs.fyi/signals/509454a6-8557-4fd7-bd6d-56e01ad15d28","signal_json_url":"https://onlylabs.fyi/signals/509454a6-8557-4fd7-bd6d-56e01ad15d28/signal.json","text":"model_released · ByteDance-Seed/Stable-DiffCoder-8B-Base · signal_desk=releases · occurred_at=2026-01-15T07:48:10+00:00 · url=https://huggingface.co/ByteDance-Seed/Stable-DiffCoder-8B-Base · hf_downloads=376 · hf_likes=17 · hf_params=8250462208 · pipeline=text-generation · license=mit"},{"ref":"E23","kind":"event","title":"ByteDance-Seed/M3-Agent-Memorization","date":"2025-08-12T11:06:59+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/M3-Agent-Memorization","signal_url":"https://onlylabs.fyi/signals/258aebdd-038d-41eb-ada0-6c2da2b3a363","signal_json_url":"https://onlylabs.fyi/signals/258aebdd-038d-41eb-ada0-6c2da2b3a363/signal.json","text":"model_released · ByteDance-Seed/M3-Agent-Memorization · signal_desk=releases · occurred_at=2025-08-12T11:06:59+00:00 · url=https://huggingface.co/ByteDance-Seed/M3-Agent-Memorization · hf_downloads=39 · hf_likes=16 · hf_params=8931813888 · license=apache-2.0"},{"ref":"E24","kind":"event","title":"ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-14B","date":"2025-10-08T07:23:59+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-14B","signal_url":"https://onlylabs.fyi/signals/dc188068-bc64-4baf-ad73-5053915ccc6b","signal_json_url":"https://onlylabs.fyi/signals/dc188068-bc64-4baf-ad73-5053915ccc6b/signal.json","text":"model_released · ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-14B · signal_desk=releases · occurred_at=2025-10-08T07:23:59+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-14B · hf_downloads=41 · hf_likes=15 · hf_params=51371520 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E25","kind":"event","title":"ByteDance-Seed/In-Place-TTT","date":"2026-04-07T05:50:45+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/In-Place-TTT","signal_url":"https://onlylabs.fyi/signals/106937f9-35dd-41e2-adde-98e74527e1b8","signal_json_url":"https://onlylabs.fyi/signals/106937f9-35dd-41e2-adde-98e74527e1b8/signal.json","text":"repo_new · ByteDance-Seed/In-Place-TTT · signal_desk=repos · occurred_at=2026-04-07T05:50:45+00:00 · url=https://github.com/ByteDance-Seed/In-Place-TTT · stars=231 · raw={\"repo\":\"ByteDance-Seed/In-Place-TTT\",\"language\":\"Python\"}"},{"ref":"E26","kind":"event","title":"ByteDance-Seed/BFS-Prover-V2-32B","date":"2025-09-30T00:05:39+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/BFS-Prover-V2-32B","signal_url":"https://onlylabs.fyi/signals/f2474015-c231-47c1-8b8b-1192859a0ff7","signal_json_url":"https://onlylabs.fyi/signals/f2474015-c231-47c1-8b8b-1192859a0ff7/signal.json","text":"model_released · ByteDance-Seed/BFS-Prover-V2-32B · signal_desk=releases · occurred_at=2025-09-30T00:05:39+00:00 · url=https://huggingface.co/ByteDance-Seed/BFS-Prover-V2-32B · hf_downloads=226 · hf_likes=13 · hf_params=32763876352 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E27","kind":"event","title":"ByteDance-Seed/Adversarial-Flow-Models","date":"2025-12-02T06:03:23+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/Adversarial-Flow-Models","signal_url":"https://onlylabs.fyi/signals/0dc8d2c5-82d2-434a-a619-f9ac64d0f827","signal_json_url":"https://onlylabs.fyi/signals/0dc8d2c5-82d2-434a-a619-f9ac64d0f827/signal.json","text":"model_released · ByteDance-Seed/Adversarial-Flow-Models · signal_desk=releases · occurred_at=2025-12-02T06:03:23+00:00 · url=https://huggingface.co/ByteDance-Seed/Adversarial-Flow-Models · hf_likes=13 · license=mit"},{"ref":"E28","kind":"event","title":"ByteDance-Seed/VINCIE-7B","date":"2026-01-06T05:59:42+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/VINCIE-7B","signal_url":"https://onlylabs.fyi/signals/65056b70-b5de-497f-9b38-280f8e0c82da","signal_json_url":"https://onlylabs.fyi/signals/65056b70-b5de-497f-9b38-280f8e0c82da/signal.json","text":"model_released · ByteDance-Seed/VINCIE-7B · signal_desk=releases · occurred_at=2026-01-06T05:59:42+00:00 · url=https://huggingface.co/ByteDance-Seed/VINCIE-7B · hf_likes=12 · license=apache-2.0"},{"ref":"E29","kind":"event","title":"ByteDance-Seed/VeOmni","date":"2025-03-28T03:42:42+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/VeOmni","signal_url":"https://onlylabs.fyi/signals/13ff5b46-cb84-4c16-8e0e-8fc8464a1442","signal_json_url":"https://onlylabs.fyi/signals/13ff5b46-cb84-4c16-8e0e-8fc8464a1442/signal.json","text":"repo_new · ByteDance-Seed/VeOmni · signal_desk=repos · occurred_at=2025-03-28T03:42:42+00:00 · url=https://github.com/ByteDance-Seed/VeOmni · stars=2004 · data_radar_lanes=Infrastructure · data_radar_terms=training, scaling · data_radar_reason=ByteDance (Doubao/Seed) has a repo signal matching infrastructure. · raw={\"repo\":\"ByteDance-Seed/VeOmni\",\"description\":\"VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo\",\"language\":\"Python\"}"},{"ref":"E30","kind":"event","title":"ByteDance-Seed/bamboo_mixer","date":"2025-08-26T00:44:02+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/bamboo_mixer","signal_url":"https://onlylabs.fyi/signals/c62b140e-da39-4806-88f1-e7db89659e5a","signal_json_url":"https://onlylabs.fyi/signals/c62b140e-da39-4806-88f1-e7db89659e5a/signal.json","text":"model_released · ByteDance-Seed/bamboo_mixer · signal_desk=releases · occurred_at=2025-08-26T00:44:02+00:00 · url=https://huggingface.co/ByteDance-Seed/bamboo_mixer · hf_likes=10 · license=cc-by-4.0"},{"ref":"E31","kind":"event","title":"ByteDance-Seed/Seed1.5-VL","date":"2025-05-11T06:00:32+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Seed1.5-VL","signal_url":"https://onlylabs.fyi/signals/ed934491-9360-4687-b39e-f10095fa9623","signal_json_url":"https://onlylabs.fyi/signals/ed934491-9360-4687-b39e-f10095fa9623/signal.json","text":"repo_new · ByteDance-Seed/Seed1.5-VL · signal_desk=repos · occurred_at=2025-05-11T06:00:32+00:00 · url=https://github.com/ByteDance-Seed/Seed1.5-VL · stars=1580 · data_radar_lanes=Evals and quality · data_radar_terms=benchmark, benchmarks · data_radar_reason=ByteDance (Doubao/Seed) has a repo signal matching evals and quality. · raw={\"repo\":\"ByteDance-Seed/Seed1.5-VL\",\"description\":\"Seed1.5-VL, a vision-language foundation model designed to advance general-purpose multimodal understanding and reasoning, achieving state-of-the-art performance on 38 out of 60 public benchmarks.\",\"language\":\"Jupyter Notebook\"}"},{"ref":"E32","kind":"event","title":"ByteDance-Seed/Triton-distributed","date":"2025-04-02T06:57:03+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Triton-distributed","signal_url":"https://onlylabs.fyi/signals/dc5f0de4-e679-4989-b8c3-77e3bd375c1b","signal_json_url":"https://onlylabs.fyi/signals/dc5f0de4-e679-4989-b8c3-77e3bd375c1b/signal.json","text":"repo_new · ByteDance-Seed/Triton-distributed · signal_desk=repos · occurred_at=2025-04-02T06:57:03+00:00 · url=https://github.com/ByteDance-Seed/Triton-distributed · stars=1457 · data_radar_lanes=Infrastructure · data_radar_terms=systems · data_radar_reason=ByteDance (Doubao/Seed) has a repo signal matching infrastructure. · raw={\"repo\":\"ByteDance-Seed/Triton-distributed\",\"description\":\"Distributed Compiler based on Triton for Parallel Systems\",\"language\":\"Python\"}"},{"ref":"E33","kind":"event","title":"ByteDance-Seed/BFS-Prover-V2-7B","date":"2025-10-06T05:22:52+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/BFS-Prover-V2-7B","signal_url":"https://onlylabs.fyi/signals/3a6e9435-66ab-438b-ac41-14daa8763419","signal_json_url":"https://onlylabs.fyi/signals/3a6e9435-66ab-438b-ac41-14daa8763419/signal.json","text":"model_released · ByteDance-Seed/BFS-Prover-V2-7B · signal_desk=releases · occurred_at=2025-10-06T05:22:52+00:00 · url=https://huggingface.co/ByteDance-Seed/BFS-Prover-V2-7B · hf_downloads=593 · hf_likes=7 · hf_params=7615616512 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E34","kind":"event","title":"ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-3B","date":"2025-10-08T07:22:45+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-3B","signal_url":"https://onlylabs.fyi/signals/c8fece10-1b61-4096-8e0e-3a9eb4f1c115","signal_json_url":"https://onlylabs.fyi/signals/c8fece10-1b61-4096-8e0e-3a9eb4f1c115/signal.json","text":"model_released · ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-3B · signal_desk=releases · occurred_at=2025-10-08T07:22:45+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-3B · hf_downloads=43 · hf_likes=7 · hf_params=11872512 · pipeline=text-generation · license=other · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E35","kind":"event","title":"ByteDance-Seed/m3-agent","date":"2025-07-30T13:12:32+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/m3-agent","signal_url":"https://onlylabs.fyi/signals/776bc08d-6253-4bfe-8e35-bfbd7fe9c3ff","signal_json_url":"https://onlylabs.fyi/signals/776bc08d-6253-4bfe-8e35-bfbd7fe9c3ff/signal.json","text":"repo_new · ByteDance-Seed/m3-agent · signal_desk=repos · occurred_at=2025-07-30T13:12:32+00:00 · url=https://github.com/ByteDance-Seed/m3-agent · stars=1379 · raw={\"repo\":\"ByteDance-Seed/m3-agent\",\"language\":\"Python\"}"},{"ref":"E36","kind":"event","title":"ByteDance-Seed/SeedVR","date":"2025-06-10T10:41:54+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/SeedVR","signal_url":"https://onlylabs.fyi/signals/66ecd70a-a89b-4779-a287-aedef9f43f7a","signal_json_url":"https://onlylabs.fyi/signals/66ecd70a-a89b-4779-a287-aedef9f43f7a/signal.json","text":"repo_new · ByteDance-Seed/SeedVR · signal_desk=repos · occurred_at=2025-06-10T10:41:54+00:00 · url=https://github.com/ByteDance-Seed/SeedVR · stars=1223 · raw={\"repo\":\"ByteDance-Seed/SeedVR\",\"description\":\"Repo for SeedVR2 (ICLR2026) & SeedVR (CVPR2025 Highlight)\",\"language\":\"Python\"}"},{"ref":"E37","kind":"event","title":"ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-14B","date":"2025-10-08T07:21:07+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-14B","signal_url":"https://onlylabs.fyi/signals/25ff0b25-e9fd-404f-a607-32955233864c","signal_json_url":"https://onlylabs.fyi/signals/25ff0b25-e9fd-404f-a607-32955233864c/signal.json","text":"model_released · ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-14B · signal_desk=releases · occurred_at=2025-10-08T07:21:07+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-14B · hf_downloads=42 · hf_likes=6 · hf_params=60960384 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E38","kind":"event","title":"ByteDance-Seed/cryofm-v2","date":"2025-12-24T03:19:32+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/cryofm-v2","signal_url":"https://onlylabs.fyi/signals/986ad3f8-f6f6-45d7-b4ed-d1b79b969c36","signal_json_url":"https://onlylabs.fyi/signals/986ad3f8-f6f6-45d7-b4ed-d1b79b969c36/signal.json","text":"model_released · ByteDance-Seed/cryofm-v2 · signal_desk=releases · occurred_at=2025-12-24T03:19:32+00:00 · url=https://huggingface.co/ByteDance-Seed/cryofm-v2 · hf_downloads=16 · hf_likes=6 · pipeline=other · license=apache-2.0"},{"ref":"E39","kind":"event","title":"ByteDance-Seed/cryofm-v1","date":"2025-12-23T15:41:52+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/cryofm-v1","signal_url":"https://onlylabs.fyi/signals/ffcb0323-1ae7-4a9a-ab38-52ddf5065b53","signal_json_url":"https://onlylabs.fyi/signals/ffcb0323-1ae7-4a9a-ab38-52ddf5065b53/signal.json","text":"model_released · ByteDance-Seed/cryofm-v1 · signal_desk=releases · occurred_at=2025-12-23T15:41:52+00:00 · url=https://huggingface.co/ByteDance-Seed/cryofm-v1 · hf_downloads=15 · hf_likes=6 · pipeline=other · license=apache-2.0"},{"ref":"E40","kind":"event","title":"ByteDance-Seed/ConfRover-interp-20M-v1.0","date":"2025-12-01T12:23:15+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/ConfRover-interp-20M-v1.0","signal_url":"https://onlylabs.fyi/signals/0c4465f9-3bd7-4228-b265-9c602877702d","signal_json_url":"https://onlylabs.fyi/signals/0c4465f9-3bd7-4228-b265-9c602877702d/signal.json","text":"model_released · ByteDance-Seed/ConfRover-interp-20M-v1.0 · signal_desk=releases · occurred_at=2025-12-01T12:23:15+00:00 · url=https://huggingface.co/ByteDance-Seed/ConfRover-interp-20M-v1.0 · hf_likes=6 · pipeline=other · license=apache-2.0"},{"ref":"E41","kind":"event","title":"ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-7B","date":"2025-10-08T07:23:28+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-7B","signal_url":"https://onlylabs.fyi/signals/41d25c88-a23b-4b10-a479-a10da104893b","signal_json_url":"https://onlylabs.fyi/signals/41d25c88-a23b-4b10-a479-a10da104893b/signal.json","text":"model_released · ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-7B · signal_desk=releases · occurred_at=2025-10-08T07:23:28+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-7B · hf_downloads=44 · hf_likes=5 · hf_params=18568256 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E42","kind":"event","title":"ByteDance-Seed/byteff2","date":"2025-08-26T02:29:56+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/byteff2","signal_url":"https://onlylabs.fyi/signals/d9458cfe-645c-4193-91a6-ac50aacdb946","signal_json_url":"https://onlylabs.fyi/signals/d9458cfe-645c-4193-91a6-ac50aacdb946/signal.json","text":"model_released · ByteDance-Seed/byteff2 · signal_desk=releases · occurred_at=2025-08-26T02:29:56+00:00 · url=https://huggingface.co/ByteDance-Seed/byteff2 · hf_likes=5 · license=cc-by-nc-4.0"},{"ref":"E43","kind":"event","title":"ByteDance-Seed/MoDA","date":"2026-03-23T06:55:29+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/MoDA","signal_url":"https://onlylabs.fyi/signals/614ce86f-192a-4207-b42c-315bffe40df6","signal_json_url":"https://onlylabs.fyi/signals/614ce86f-192a-4207-b42c-315bffe40df6/signal.json","text":"repo_new · ByteDance-Seed/MoDA · signal_desk=repos · occurred_at=2026-03-23T06:55:29+00:00 · url=https://github.com/ByteDance-Seed/MoDA · stars=7 · raw={\"repo\":\"ByteDance-Seed/MoDA\",\"language\":\"Python\"}"},{"ref":"E44","kind":"event","title":"ByteDance-Seed/seed-oss","date":"2025-08-08T00:48:59+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/seed-oss","signal_url":"https://onlylabs.fyi/signals/9bc27772-7273-498c-8961-de04f9edd791","signal_json_url":"https://onlylabs.fyi/signals/9bc27772-7273-498c-8961-de04f9edd791/signal.json","text":"repo_new · ByteDance-Seed/seed-oss · signal_desk=repos · occurred_at=2025-08-08T00:48:59+00:00 · url=https://github.com/ByteDance-Seed/seed-oss · stars=885 · raw={\"repo\":\"ByteDance-Seed/seed-oss\",\"language\":\"Python\"}"},{"ref":"E45","kind":"event","title":"ByteDance-Seed/Seed-Thinking-v1.5","date":"2025-04-10T10:05:49+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Seed-Thinking-v1.5","signal_url":"https://onlylabs.fyi/signals/030c0f27-4a50-4aec-ba03-c4d21e507b49","signal_json_url":"https://onlylabs.fyi/signals/030c0f27-4a50-4aec-ba03-c4d21e507b49/signal.json","text":"repo_new · ByteDance-Seed/Seed-Thinking-v1.5 · signal_desk=repos · occurred_at=2025-04-10T10:05:49+00:00 · url=https://github.com/ByteDance-Seed/Seed-Thinking-v1.5 · stars=812 · raw={\"repo\":\"ByteDance-Seed/Seed-Thinking-v1.5\"}"},{"ref":"E46","kind":"event","title":"ByteDance-Seed/ConfRover-base-20M-v1.0","date":"2025-12-01T11:41:17+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/ConfRover-base-20M-v1.0","signal_url":"https://onlylabs.fyi/signals/1853e25a-29e6-4ff4-a5af-44e5a7a59e31","signal_json_url":"https://onlylabs.fyi/signals/1853e25a-29e6-4ff4-a5af-44e5a7a59e31/signal.json","text":"model_released · ByteDance-Seed/ConfRover-base-20M-v1.0 · signal_desk=releases · occurred_at=2025-12-01T11:41:17+00:00 · url=https://huggingface.co/ByteDance-Seed/ConfRover-base-20M-v1.0 · hf_likes=4 · license=apache-2.0"},{"ref":"E47","kind":"event","title":"ByteDance-Seed/VideoWorld","date":"2025-01-15T06:49:23+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/VideoWorld","signal_url":"https://onlylabs.fyi/signals/6c67c08a-c31a-47fa-9496-90cd1152ba5f","signal_json_url":"https://onlylabs.fyi/signals/6c67c08a-c31a-47fa-9496-90cd1152ba5f/signal.json","text":"repo_new · ByteDance-Seed/VideoWorld · signal_desk=repos · occurred_at=2025-01-15T06:49:23+00:00 · url=https://github.com/ByteDance-Seed/VideoWorld · stars=790 · data_radar_lanes=Data demand · data_radar_terms=label · data_radar_reason=ByteDance (Doubao/Seed) has a repo signal matching data demand. · raw={\"repo\":\"ByteDance-Seed/VideoWorld\",\"description\":\"[CVPR 2025] VideoWorld is a simple generative model that learns purely from unlabeled videos—much like how babies learn by observing their environment.\",\"language\":\"Python\"}"},{"ref":"E48","kind":"event","title":"ByteDance-Seed/Seed-Coder","date":"2025-04-21T06:52:36+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Seed-Coder","signal_url":"https://onlylabs.fyi/signals/650d9bc3-795e-433b-8462-a411f572bf8f","signal_json_url":"https://onlylabs.fyi/signals/650d9bc3-795e-433b-8462-a411f572bf8f/signal.json","text":"repo_new · ByteDance-Seed/Seed-Coder · signal_desk=repos · occurred_at=2025-04-21T06:52:36+00:00 · url=https://github.com/ByteDance-Seed/Seed-Coder · stars=754 · raw={\"repo\":\"ByteDance-Seed/Seed-Coder\",\"description\":\"Seed-Coder is a family of lightweight open-source code LLMs comprising base, instruct and reasoning models, developed by ByteDance Seed.\"}"},{"ref":"E49","kind":"event","title":"ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-14B","date":"2025-10-08T07:20:11+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-14B","signal_url":"https://onlylabs.fyi/signals/f7fa636f-3386-4c02-8037-99771e71adb1","signal_json_url":"https://onlylabs.fyi/signals/f7fa636f-3386-4c02-8037-99771e71adb1/signal.json","text":"model_released · ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-14B · signal_desk=releases · occurred_at=2025-10-08T07:20:11+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-14B · hf_downloads=42 · hf_likes=3 · hf_params=51124224 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E50","kind":"event","title":"ByteDance-Seed/TraceAnything","date":"2025-10-15T05:59:37+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/TraceAnything","signal_url":"https://onlylabs.fyi/signals/4460d2f3-e801-4786-b00e-1ebd433c4422","signal_json_url":"https://onlylabs.fyi/signals/4460d2f3-e801-4786-b00e-1ebd433c4422/signal.json","text":"repo_new · ByteDance-Seed/TraceAnything · signal_desk=repos · occurred_at=2025-10-15T05:59:37+00:00 · url=https://github.com/ByteDance-Seed/TraceAnything · stars=538 · raw={\"repo\":\"ByteDance-Seed/TraceAnything\",\"description\":\"[ICLR 2026] Trace Anything: Representing Any Video in 4D via Trajectory Fields\",\"language\":\"Python\"}"},{"ref":"E51","kind":"event","title":"ByteDance-Seed/Seed-Prover","date":"2025-07-16T07:37:21+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Seed-Prover","signal_url":"https://onlylabs.fyi/signals/b8ca1839-0a15-4666-81e9-ab9ee414c0cf","signal_json_url":"https://onlylabs.fyi/signals/b8ca1839-0a15-4666-81e9-ab9ee414c0cf/signal.json","text":"repo_new · ByteDance-Seed/Seed-Prover · signal_desk=repos · occurred_at=2025-07-16T07:37:21+00:00 · url=https://github.com/ByteDance-Seed/Seed-Prover · stars=433 · raw={\"repo\":\"ByteDance-Seed/Seed-Prover\",\"language\":\"Lean\"}"},{"ref":"E52","kind":"event","title":"ByteDance-Seed/EvaLearn","date":"2025-06-03T04:41:27+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/EvaLearn","signal_url":"https://onlylabs.fyi/signals/af1984e8-e741-487e-a021-e0a3eb7893da","signal_json_url":"https://onlylabs.fyi/signals/af1984e8-e741-487e-a021-e0a3eb7893da/signal.json","text":"repo_new · ByteDance-Seed/EvaLearn · signal_desk=repos · occurred_at=2025-06-03T04:41:27+00:00 · url=https://github.com/ByteDance-Seed/EvaLearn · stars=431 · data_radar_lanes=Evals and quality · data_radar_terms=eval, benchmark · data_radar_reason=ByteDance (Doubao/Seed) has a repo signal matching evals and quality. · raw={\"repo\":\"ByteDance-Seed/EvaLearn\",\"description\":\"EvaLearn is a pioneering benchmark designed to evaluate large language models (LLMs) on their learning capability and efficiency in challenging tasks.\",\"language\":\"Python\"}"},{"ref":"E53","kind":"event","title":"ByteDance-Seed/DAComp","date":"2025-12-08T10:08:00+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/DAComp","signal_url":"https://onlylabs.fyi/signals/c4d6c5e6-066e-4e64-ac20-c60a236990f3","signal_json_url":"https://onlylabs.fyi/signals/c4d6c5e6-066e-4e64-ac20-c60a236990f3/signal.json","text":"repo_new · ByteDance-Seed/DAComp · signal_desk=repos · occurred_at=2025-12-08T10:08:00+00:00 · url=https://github.com/ByteDance-Seed/DAComp · stars=425 · data_radar_lanes=Data demand, Evals and quality · data_radar_terms=data, benchmark · data_radar_reason=ByteDance (Doubao/Seed) has a repo signal matching data demand, evals and quality. · raw={\"repo\":\"ByteDance-Seed/DAComp\",\"description\":\"[ICLR 2026] DAComp: Benchmarking Data Agents across the Full Data Intelligence Lifecycle\",\"language\":\"Python\"}"},{"ref":"E54","kind":"event","title":"ByteDance-Seed/ShadowKV","date":"2024-10-22T02:32:21+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/ShadowKV","signal_url":"https://onlylabs.fyi/signals/6d0dad3c-68d2-4c02-9211-3099d52c0cef","signal_json_url":"https://onlylabs.fyi/signals/6d0dad3c-68d2-4c02-9211-3099d52c0cef/signal.json","text":"repo_new · ByteDance-Seed/ShadowKV · signal_desk=repos · occurred_at=2024-10-22T02:32:21+00:00 · url=https://github.com/ByteDance-Seed/ShadowKV · stars=306 · data_radar_lanes=Infrastructure · data_radar_terms=inference · data_radar_reason=ByteDance (Doubao/Seed) has a repo signal matching infrastructure. · raw={\"repo\":\"ByteDance-Seed/ShadowKV\",\"description\":\"[ICML 2025 Spotlight] ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference\",\"language\":\"Python\"}"},{"ref":"E55","kind":"event","title":"ByteDance-Seed/ByteCheckpoint","date":"2025-03-20T04:03:05+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/ByteCheckpoint","signal_url":"https://onlylabs.fyi/signals/ebbf395f-e4b0-4196-a9dc-606ead0a5b8f","signal_json_url":"https://onlylabs.fyi/signals/ebbf395f-e4b0-4196-a9dc-606ead0a5b8f/signal.json","text":"repo_new · ByteDance-Seed/ByteCheckpoint · signal_desk=repos · occurred_at=2025-03-20T04:03:05+00:00 · url=https://github.com/ByteDance-Seed/ByteCheckpoint · stars=282 · raw={\"repo\":\"ByteDance-Seed/ByteCheckpoint\",\"description\":\"ByteCheckpoint: An Unified Checkpointing Library for LFMs\",\"language\":\"Python\"}"},{"ref":"E56","kind":"event","title":"ByteDance-Seed/Seed-1.8","date":"2025-12-17T12:12:48+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/Seed-1.8","signal_url":"https://onlylabs.fyi/signals/24c56115-3659-4a88-a220-b010f25ccde0","signal_json_url":"https://onlylabs.fyi/signals/24c56115-3659-4a88-a220-b010f25ccde0/signal.json","text":"repo_new · ByteDance-Seed/Seed-1.8 · signal_desk=repos · occurred_at=2025-12-17T12:12:48+00:00 · url=https://github.com/ByteDance-Seed/Seed-1.8 · stars=218 · raw={\"repo\":\"ByteDance-Seed/Seed-1.8\",\"language\":\"Jupyter Notebook\"}"},{"ref":"E57","kind":"event","title":"ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-3B","date":"2025-10-08T07:20:34+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-3B","signal_url":"https://onlylabs.fyi/signals/832efb71-50c8-4125-ae7e-d016f4c4370c","signal_json_url":"https://onlylabs.fyi/signals/832efb71-50c8-4125-ae7e-d016f4c4370c/signal.json","text":"model_released · ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-3B · signal_desk=releases · occurred_at=2025-10-08T07:20:34+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-3B · hf_downloads=46 · hf_likes=1 · hf_params=12982464 · pipeline=text-generation · license=other · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E58","kind":"event","title":"ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-7B","date":"2025-10-08T07:19:54+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-7B","signal_url":"https://onlylabs.fyi/signals/64755bb0-d0b2-4c36-af5d-dd7529c509c1","signal_json_url":"https://onlylabs.fyi/signals/64755bb0-d0b2-4c36-af5d-dd7529c509c1/signal.json","text":"model_released · ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-7B · signal_desk=releases · occurred_at=2025-10-08T07:19:54+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-7B · hf_downloads=41 · hf_likes=1 · hf_params=18468352 · pipeline=text-generation · license=apache-2.0 · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E59","kind":"event","title":"ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-3B","date":"2025-10-08T07:18:51+00:00","date_source":"source","source_url":"https://huggingface.co/ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-3B","signal_url":"https://onlylabs.fyi/signals/42e0eff3-d624-4e92-80c1-f069144c6a00","signal_json_url":"https://onlylabs.fyi/signals/42e0eff3-d624-4e92-80c1-f069144c6a00/signal.json","text":"model_released · ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-3B · signal_desk=releases · occurred_at=2025-10-08T07:18:51+00:00 · url=https://huggingface.co/ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-3B · hf_downloads=37 · hf_likes=1 · hf_params=11801088 · pipeline=text-generation · license=other · raw={\"derived_reason\":\"first-party-finetune\"}"},{"ref":"E60","kind":"event","title":"ByteDance-Seed/triton","date":"2025-08-28T11:41:54+00:00","date_source":"source","source_url":"https://github.com/ByteDance-Seed/triton","signal_url":"https://onlylabs.fyi/signals/d2b2cbf4-7ee7-443d-9695-e010c155d8e6","signal_json_url":"https://onlylabs.fyi/signals/d2b2cbf4-7ee7-443d-9695-e010c155d8e6/signal.json","text":"repo_forked · ByteDance-Seed/triton · signal_desk=forks · occurred_at=2025-08-28T11:41:54+00:00 · url=https://github.com/ByteDance-Seed/triton · raw={\"repo\":\"ByteDance-Seed/triton\",\"parent\":\"triton-lang/triton\"}"}]}