ViperEkura b89f8436ea refactor: 将KV缓存槽位映射下沉到模型注意力层，移除_remap_kv和_writeback_kv

2026-05-06 20:01:22 +08:00

..

config

feat: 实现模型动态注册机制

2026-04-05 19:38:12 +08:00

dataset

chore: 修改错误拼写

2026-04-06 10:37:19 +08:00

inference

refactor: 将KV缓存槽位映射下沉到模型注意力层，移除_remap_kv和_writeback_kv

2026-05-06 20:01:22 +08:00

model

refactor: 将KV缓存槽位映射下沉到模型注意力层，移除_remap_kv和_writeback_kv

2026-05-06 20:01:22 +08:00

parallel

refactor: 优化参数传递，清理导入样式

2026-04-03 22:06:32 +08:00

tokenize

refactor: 精简推理引擎代码，优化参数传递规范

2026-04-09 14:17:48 +08:00

trainer

chore: 修改拼写错误问题

2026-04-06 09:28:16 +08:00

__init__.py

chore: 优化未使用的模块

2026-04-06 09:54:17 +08:00

factory.py

chore: 修改类名，优化导入顺序

2026-04-05 22:27:57 +08:00

serialization.py

refactor: 从data 模块分离tokenizer

2026-04-04 16:12:58 +08:00