Huanshere / Huanshere/VideoLingo
🙋♂️【求助】使用 OpenAI API KEY + Replicate Cloud 翻译 1 小时 11 分钟的视频频频失败
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 18.5k
- Forks
- 2k
- Avg merge
- 7h 41m
- Merged PRs (30d)
- 12
Description
背景:
我正在使用 videoling 项目通过以下配置方式尝试翻译一段 1 小时 11 分钟的视频字幕:
LLM 模型使用: OpenAI API(目前为 GPT-4.1)
推理平台: Replicate Cloud
视频长度: 1小时11分钟
容器环境: Docker Desktop (个人版)
问题描述:
在执行翻译过程中,经常遇到以下错误并导致流程中断:
ValueError: Empty translation rows detected, this is likely due to weak model, please use Claude-3-5-sonnet or manually adjust the empty rows.
但我觉得可能并不是模型的问题?我选用的是 GPT4.1
完整报错截图如下:
图片 1
图片 2
从日志来看,翻译过程中会生成空行数据,程序随后检测到 translation_results.xlsx 和 translation_results_for_subtitles.xlsx 中存在空行并直接抛出异常,导致后续步骤无法继续执行。
❓我想了解:
-
是否是因为视频中存在一段时间的空白或停顿,模型误判导致跳过翻译?
-
目前 GPT-4o 的表现是否确实不如 Claude-3.5-sonnet 在处理较长视频翻译任务?
-
是否有推荐的自动 retry 或 fallback 策略?
-
我该如何配置以规避这个问题?
💻环境信息:
-
容器镜像: rqlove/videoling:latest
-
Docker Desktop 状态: Engine running,资源占用正常
-
报错发生时间: 2025-07-28 15:47(UTC+8)
期待您们的帮助,非常感谢!🙏
如需我提供更多日志或配置详情,请随时告知!
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the reported logs and inspect translation_results.xlsx and translation_results_for_subtitles.xlsx for the empty rows associated with the failed run. Check the translation pipeline entry points that generate and validate these files; done means identifying a reproducible cause and defining a verified retry, fallback, or configuration fix.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- docker, python
- Domain
- ai, backend, localization
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100