Logic
|
ff7c9c1f2a
|
feat: add vision transfer backbones and IMF variants
|
2026-04-09 14:02:24 +08:00 |
|
Logic
|
d84bc6876e
|
feat(vla): align transformer training stack and rollout validation
|
2026-03-31 15:39:20 +08:00 |
|
gouhanke
|
7d39933a5b
|
feat: 缓存worker内的句柄
|
2026-03-04 10:49:41 +08:00 |
|
gouhanke
|
926d8cf894
|
chore: 加载时将图像缩放到224*224， resnet禁用crop
|
2026-02-12 15:02:18 +08:00 |
|
gouhanke
|
83cd55e67b
|
添加pad_loss
|
2026-02-11 20:33:26 +08:00 |
|
gouhanke
|
b42c1c68fd
|
debug: 将归一化放在GPU上
|
2026-02-11 17:13:55 +08:00 |
|
gouhanke
|
320369ffb8
|
debug: 归一化图像到[0, 1]
|
2026-02-11 16:47:39 +08:00 |
|
gouhanke
|
130d4bb3c5
|
refactor：大重构
|
2026-02-11 15:53:55 +08:00 |
|
gouhanke
|
88b9c10a75
|
refactor(dataset): 重新创建robotdataset最小实现
- 内部实现__getitem__参数，可以通过滑动窗口进行采样
-
|
2026-02-10 10:26:19 +08:00 |
|
gouhanke
|
a43a2e3d18
|
chore: 删除多余脚本
|
2026-02-06 13:45:35 +08:00 |
|
gouhanke
|
31419a6fc1
|
chore(camera): 添加front相机
|
2026-02-06 11:53:01 +08:00 |
|
gouhanke
|
66009473ad
|
debug(inference): 添加推理阶段qpos归一化
|
2026-02-06 09:00:44 +08:00 |
|
gouhanke
|
b0a944f7aa
|
feat(train): 跑通训练脚本
|
2026-02-05 14:08:43 +08:00 |
|
gouhanke
|
dd2749cb12
|
feat: 更新框架，新增数据及定义和backbone
|
2026-02-05 01:44:43 +08:00 |
|
gouhanke
|
f5e2eca809
|
debug(train): 在siglip和DiffusionHead下跑通训练流程
|
2026-02-03 17:42:32 +08:00 |
|
gouhanke
|
3b58760469
|
跑通配置和训练脚本
|
2026-02-03 16:51:04 +08:00 |
|
gouhanke
|
d3863ea1dd
|
feat(dataset): 定义VLAChunkedDataset类，构建数据可视化工具
|
2026-02-03 15:24:09 +08:00 |
|
gouhanke
|
57acfd645f
|
feat(vla): vla框架初始化
|
2026-02-03 14:18:30 +08:00 |
|