data-loader/fivevla-posttrain-caches
FiveVLA post-train 缓存迁移包 这是本项目实际使用的 34 份 FiveVLA 样本缓存，源文件共约 410 MB，压缩包约 20.6 MB。 保留样本编号、顺序、bucket、训练/验证划分；原服务器路径已转换成占位符， 恢复时自动写入新路径，并调整 NumPy 字符串容量以防路径截断。 不含原始图片、驾驶标签、模型权重，也不含 bucket 定义索引。 恢复位置（相对于 FiveVLA 子项目） 用途 buckets_cache/debug-False/hist-1/*.npy 第一阶段，16 份训练 bucket + 1 份验证缓存 buckets_cache/debug-False/hist-4/*.npy 第二阶段 RAM，16 份训练 bucket + 1 份验证缓存 本快照对应 SimLingo expert 配方：dreamer=False、split_seed=98765、路线级 99/1 划分。 dreamer-False 是文件名的布尔标记；这些缓存不包含 Dreamer… See the full description on the dataset page: https://huggingface.co/datasets/data-loader/fivevla-posttrain-caches.
| Author / organization | data-loader |
|---|---|
| License (as declared) | other |
| Downloads | 0 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 21 October 2026 |
| Last updated at source | 21 October 2026 |
More from data-loader
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 21 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.