eCity-Omni is an open-weight model from Supconit. Its published files total 21.3 GB.
基于 Qwen2 的全模态（视频 + 音频 + 文本）大模型，支持视频时序定位（temporal grounding）与多模态开放式问答。本文档只涵盖模型调用到结果预测的推理流程。 另外需要系统装有 ffmpeg（用于从视频提取音频）。 下载 eCityOmini checkpoint: https://huggingface.co/mxxxxxxxxxxxxxxxxx/ChronusOmni checkpoint 目录内需包含 LLM 权重、tokenizer、config 以及 BEATs…
Parameters—
Context—
Weights21.2 GB
License—
AccessOpen weights
Monthly Downloads—
Model Card
基于 Qwen2 的全模态（视频 + 音频 + 文本）大模型，支持视频时序定位（temporal grounding）与多模态开放式问答。本文档只涵盖模型调用到结果预测的推理流程。 另外需要系统装有 ffmpeg（用于从视频提取音频）。 下载 eCityOmini checkpoint: https://huggingface.co/mxxxxxxxxxxxxxxxxx/ChronusOmni checkpoint 目录内需包含 LLM 权重、tokenizer、config 以及 BEATs 语音编码器权重（BEATsiter3plusAS2MfinetunedonAS2Mcpt2.pt）。视觉塔权重已包含在 LLM checkpoint 中，无需单独下载。 输入视频路径 + 问题，一条命令完成 ffmpeg 提取音频 → 均匀抽帧 + 时间戳 → 模型推理 → 打印结果，临时音频文件自动清理： 模型按 From second{X} to second{Y}. 回答，脚本解析后打印： 即 [start, end]，保留一位小数；解析失败时回退打印模型原始输出。 无外网服务器报错 couldn't connect to huggingface.co... openai/clip-vit-large-patch14 加载视觉塔时 chronus/model/multimodalencoder/oryxvit.py 会从 HuggingFace 下载 CLIP image processor 的配置文件（几 KB）。任选一种解决： 1. 运行前设置镜像：export…
Excerpt from the card by Supconit.
Identity and Version
- Repository
- eCityAI-Sup/eCity-Omni
- Publisher
- Supconit
- Task
- Not stated by the source
- Modality
- Other
- Library
- Not stated by the source
- Parameters
- Not stated by the source
- Languages
- Not stated by the source
- Revision
- ae326d7bfc3558fc29181a3653969a84e687c166
- First published
- 2026-10-16
- Last updated
- 2026-10-16
Files and Weights
151 files, 21.3 GB in total. The weights are 12 files totalling 21.2 GB in bin, npz, pt, pth.
Weights12 files · 21.2 GB
Configuration88 files · 59.8 MB
Tokenizer8 files · 6.7 MB
Documentation2 files · 3.9 KB
Other37 files · 1.4 MB
Repository4 files · 9.4 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| ChronusOmni/BEATs_iter3_plus_AS2M_finetuned_on_AS2M_cpt2.pt | Weights | 363.1 MB | e5815275a04b |
| ChronusOmni/large-v3.pt | Weights | 3.1 GB | e5b1a55b89c1 |
| ChronusOmni/pytorch_model-00001-of-00004.bin | Weights | 4.9 GB | 057e84f91fd1 |
| ChronusOmni/pytorch_model-00002-of-00004.bin | Weights | 4.9 GB | 253b847e8677 |
| ChronusOmni/pytorch_model-00003-of-00004.bin | Weights | 5.0 GB | 9368d77031f7 |
| ChronusOmni/pytorch_model-00004-of-00004.bin | Weights | 3.0 GB | 99579f9cea9f |
| ChronusOmni/rng_state_0.pth | Weights | 15.0 KB | 035c0380f12c |
| ChronusOmni/rng_state_1.pth | Weights | 15.0 KB | ecf5f2c204bf |
| ChronusOmni/rng_state_2.pth | Weights | 15.0 KB | 20e5b777de5d |
| ChronusOmni/rng_state_3.pth | Weights | 15.0 KB | 1cd44b8a0144 |
| ChronusOmni/training_args.bin | Weights | 8.5 KB | 7b62d715f309 |
| chronus/model/speech_encoder/whisper/assets/mel_filters.npz | Weights | 4.3 KB | 7450ae70723a |
| ChronusOmni/added_tokens.json | Configuration | 735 B | — |
| ChronusOmni/config.json | Configuration | 1.8 KB | — |
| ChronusOmni/generation_config.json | Configuration | 289 B | — |
| ChronusOmni/pytorch_model.bin.index.json | Configuration | 140.0 KB | — |
| ChronusOmni/special_tokens_map.json | Configuration | 610 B | — |
| ChronusOmni/trainer_state.json | Configuration | 456.1 KB | — |
| ChronusOmni/zero_to_fp32.py | Configuration | 33.1 KB | — |
| chronus/arguments.py | Configuration | 2.2 KB | — |
| chronus/constants.py | Configuration | 351 B | — |
| chronus/conversation.py | Configuration | 8.6 KB | — |
| chronus/datasets/preprocess.py | Configuration | 8.8 KB | — |
| chronus/mm_utils.py | Configuration | 9.8 KB | — |
| chronus/model/__init__.py | Configuration | 82 B | — |
| chronus/model/builder.py | Configuration | 4.1 KB | — |
| chronus/model/chronus_arch.py | Configuration | 23.1 KB | — |
| chronus/model/language_model/chronus_qwen.py | Configuration | 8.9 KB | — |
| chronus/model/multimodal_encoder/builder.py | Configuration | 481 B | — |
| chronus/model/multimodal_encoder/oryx_vit.py | Configuration | 37.8 KB | — |
| chronus/model/multimodal_projector/builder.py | Configuration | 5.7 KB | — |
| chronus/model/multimodal_projector/pooler_projector.py | Configuration | 2.2 KB | — |
| chronus/model/multimodal_resampler/builder.py | Configuration | 519 B | — |
| chronus/model/speech_encoder/beats/BEATs.py | Configuration | 6.9 KB | — |
| chronus/model/speech_encoder/beats/__init__.py | Configuration | — | |
| chronus/model/speech_encoder/beats/backbone.py | Configuration | 30.8 KB | — |
| chronus/model/speech_encoder/beats/kaldi.py | Configuration | 36.8 KB | — |
| chronus/model/speech_encoder/beats/modules.py | Configuration | 7.4 KB | — |
| chronus/model/speech_encoder/beats/quantizer.py | Configuration | 8.2 KB | — |
| chronus/model/speech_encoder/builder.py | Configuration | 531 B | — |
| chronus/model/speech_encoder/speech_encoder.py | Configuration | 3.5 KB | — |
| chronus/model/speech_encoder/whisper/__init__.py | Configuration | 7.4 KB | — |
| chronus/model/speech_encoder/whisper/__main__.py | Configuration | 35 B | — |
| chronus/model/speech_encoder/whisper/audio.py | Configuration | 4.9 KB | — |
| chronus/model/speech_encoder/whisper/decoding.py | Configuration | 32.2 KB | — |
| chronus/model/speech_encoder/whisper/model.py | Configuration | 11.8 KB | — |
| chronus/model/speech_encoder/whisper/normalizers/__init__.py | Configuration | 130 B | — |
| chronus/model/speech_encoder/whisper/normalizers/basic.py | Configuration | 2.1 KB | — |
| chronus/model/speech_encoder/whisper/normalizers/english.json | Configuration | 56.1 KB | — |
| chronus/model/speech_encoder/whisper/normalizers/english.py | Configuration | 20.9 KB | — |
| chronus/model/speech_encoder/whisper/timing.py | Configuration | 12.7 KB | — |
| chronus/model/speech_encoder/whisper/transcribe.py | Configuration | 30.4 KB | — |
| chronus/model/speech_encoder/whisper/triton_ops.py | Configuration | 3.6 KB | — |
| chronus/model/speech_encoder/whisper/utils.py | Configuration | 11.5 KB | — |
| chronus/model/speech_encoder/whisper/version.py | Configuration | 25 B | — |
| chronus/model/speech_projector/builder.py | Configuration | 366 B | — |
| chronus/model/speech_projector/speech_projector.py | Configuration | 1.3 KB | — |
| chronus/serve/controller.py | Configuration | 10.0 KB | — |
| chronus/serve/gradio_web_server.py | Configuration | 11.9 KB | — |
| chronus/serve/model_worker.py | Configuration | 11.0 KB | — |
| chronus/train/chronus_trainer.py | Configuration | 21.2 KB | — |
| chronus/train/chronus_trainer_grpo.py | Configuration | 36.4 KB | — |
| chronus/train/train.py | Configuration | 55.7 KB | — |
| chronus/train/train_grpo.py | Configuration | 10.3 KB | — |
| chronus/utils.py | Configuration | 7.5 KB | — |
| data/test/a2t_openqa.json | Configuration | 3.0 MB | — |
| data/test/a2v_openqa.json | Configuration | 3.1 MB | — |
| data/test/longvale_test_captioning.json | Configuration | 1.1 MB | — |
| data/test/longvale_test_grounding.json | Configuration | 17.3 MB | 187c0f8267a3 |
| data/test/longvale_test_seg_captioning.json | Configuration | 13.2 MB | b2d825ec3531 |
| data/test/t2a_openqa.json | Configuration | 1.9 MB | — |
| data/test/t2v_openqa.json | Configuration | 1.5 MB | — |
| data/test/v2a_openqa.json | Configuration | 2.7 MB | — |
| data/test/v2t_openqa.json | Configuration | 2.2 MB | — |
| data/vtg_data/test/VTG_dataset.json | Configuration | 3.9 MB | — |
| data/vtg_data/test/VTG_dataset_fixed.json | Configuration | 4.1 MB | — |
| data/vtg_data/test/VTG_dataset_split1.json | Configuration | 267.1 KB | — |
| data/vtg_data/test/VTG_dataset_split10.json | Configuration | 265.3 KB | — |
| data/vtg_data/test/VTG_dataset_split11.json | Configuration | 266.5 KB | — |
| data/vtg_data/test/VTG_dataset_split12.json | Configuration | 265.2 KB | — |
| data/vtg_data/test/VTG_dataset_split13.json | Configuration | 269.5 KB | — |
| data/vtg_data/test/VTG_dataset_split14.json | Configuration | 268.5 KB | — |
| data/vtg_data/test/VTG_dataset_split15.json | Configuration | 266.3 KB | — |
| data/vtg_data/test/VTG_dataset_split16.json | Configuration | 263.2 KB | — |
| data/vtg_data/test/VTG_dataset_split17.json | Configuration | 261.8 KB | — |
| data/vtg_data/test/VTG_dataset_split2.json | Configuration | 265.7 KB | — |
| data/vtg_data/test/VTG_dataset_split3.json | Configuration | 267.9 KB | — |
| data/vtg_data/test/VTG_dataset_split4.json | Configuration | 266.6 KB | — |
| data/vtg_data/test/VTG_dataset_split5.json | Configuration | 264.3 KB | — |
| data/vtg_data/test/VTG_dataset_split6.json | Configuration | 265.4 KB | — |
| data/vtg_data/test/VTG_dataset_split7.json | Configuration | 265.3 KB | — |
| data/vtg_data/test/VTG_dataset_split8.json | Configuration | 267.5 KB | — |
| data/vtg_data/test/VTG_dataset_split9.json | Configuration | 268.9 KB | — |
| extract_audio.py | Configuration | 2.8 KB | — |
| multimodel.py | Configuration | 9.5 KB | — |
| scripts/ds_config_zero3_offload_grpo.json | Configuration | 1.3 KB | — |
| scripts/zero2.json | Configuration | 556 B | — |
| scripts/zero2_offload.json | Configuration | 1.1 KB | — |
| scripts/zero3.json | Configuration | 801 B | — |
| scripts/zero3_offload.json | Configuration | 1.3 KB | — |
| ChronusOmni/README.md | Documentation | 33 B | — |
| README.md | Documentation | 3.9 KB | — |
| ChronusOmni/latest | Other | 15 B | — |
| chronus/__pycache__/arguments.cpython-310.pyc | Other | 2.7 KB | — |
| chronus/__pycache__/constants.cpython-310.pyc | Other | 518 B | — |
| chronus/__pycache__/conversation.cpython-310.pyc | Other | 6.2 KB | — |
| chronus/__pycache__/mm_utils.cpython-310.pyc | Other | 6.7 KB | — |
| chronus/__pycache__/utils.cpython-310.pyc | Other | 7.5 KB | — |
| chronus/datasets/__pycache__/preprocess.cpython-310.pyc | Other | 7.3 KB | — |
| chronus/model/__pycache__/__init__.cpython-310.pyc | Other | 244 B | — |
| chronus/model/__pycache__/builder.cpython-310.pyc | Other | 3.3 KB | — |
| chronus/model/__pycache__/chronus_arch.cpython-310.pyc | Other | 13.4 KB | — |
| chronus/model/language_model/__pycache__/chronus_qwen.cpython-310.pyc | Other | 5.5 KB | — |
| chronus/model/multimodal_encoder/__pycache__/builder.cpython-310.pyc | Other | 589 B | — |
| chronus/model/multimodal_encoder/__pycache__/oryx_vit.cpython-310.pyc | Other | 27.9 KB | — |
| chronus/model/multimodal_projector/__pycache__/builder.cpython-310.pyc | Other | 4.4 KB | — |
| chronus/model/multimodal_projector/__pycache__/pooler_projector.cpython-310.pyc | Other | 2.6 KB | — |
| chronus/model/multimodal_resampler/__pycache__/builder.cpython-310.pyc | Other | 1.1 KB | — |
| chronus/model/speech_encoder/__pycache__/builder.cpython-310.pyc | Other | 606 B | — |
| chronus/model/speech_encoder/__pycache__/speech_encoder.cpython-310.pyc | Other | 3.2 KB | — |
| chronus/model/speech_encoder/beats/__pycache__/BEATs.cpython-310.pyc | Other | 4.1 KB | — |
| chronus/model/speech_encoder/beats/__pycache__/__init__.cpython-310.pyc | Other | 163 B | — |
| chronus/model/speech_encoder/beats/__pycache__/backbone.cpython-310.pyc | Other | 16.8 KB | — |
| chronus/model/speech_encoder/beats/__pycache__/kaldi.cpython-310.pyc | Other | 26.5 KB | — |
| chronus/model/speech_encoder/beats/__pycache__/modules.cpython-310.pyc | Other | 6.2 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/__init__.cpython-310.pyc | Other | 6.4 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/audio.cpython-310.pyc | Other | 4.6 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/decoding.cpython-310.pyc | Other | 26.1 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/model.cpython-310.pyc | Other | 11.3 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/timing.cpython-310.pyc | Other | 9.7 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/transcribe.cpython-310.pyc | Other | 20.0 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/utils.cpython-310.pyc | Other | 10.2 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/version.cpython-310.pyc | Other | 188 B | — |
| chronus/model/speech_projector/__pycache__/builder.cpython-310.pyc | Other | 505 B | — |
| chronus/model/speech_projector/__pycache__/speech_projector.cpython-310.pyc | Other | 1.4 KB | — |
| data/vtg_data/VTG_QA.jsonl | Other | 1.1 MB | — |
| requirements.txt | Other | 3.9 KB | — |
| scripts/finetune_ola.sh | Other | 2.1 KB | — |
| scripts/train_grpo.sh | Other | 1.9 KB | — |
| .gitattributes | Repository | 1.7 KB | — |
| .gitignore | Repository | 39 B | — |
| ChronusOmni/.gitattributes | Repository | 1.5 KB | — |
| scripts/.DS_Store | Repository | 6.1 KB | — |
| ChronusOmni/merges.txt | Tokenizer | 1.7 MB | — |
| ChronusOmni/tokenizer_config.json | Tokenizer | 8.3 KB | — |
| ChronusOmni/vocab.json | Tokenizer | 3.4 MB | — |
| chronus/model/speech_encoder/beats/Tokenizers.py | Tokenizer | 6.5 KB | — |
| chronus/model/speech_encoder/whisper/__pycache__/tokenizer.cpython-310.pyc | Tokenizer | 12.1 KB | — |
| chronus/model/speech_encoder/whisper/assets/gpt2.tiktoken | Tokenizer | 835.6 KB | — |
| chronus/model/speech_encoder/whisper/assets/multilingual.tiktoken | Tokenizer | 816.7 KB | — |
| chronus/model/speech_encoder/whisper/tokenizer.py | Tokenizer | 12.3 KB | — |
License and Download
- License
- Not stated by the source
- Access
- Open weights, no gate
- Download size
- 21.2 GB
Download from Supconit
Released by Supconit through its official repository on Hugging Face.
Built From
- Described by arXiv:2512.09841
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 21.2 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.