|
conftest.py
|
feat(model): 添加QK归一化和门控注意力支持
|
2026-01-05 16:14:44 +08:00 |
|
test_callbacks.py
|
refactor(trainer): 优化trainer 结构
|
2025-12-07 21:23:05 +08:00 |
|
test_checkpoint.py
|
fix(trainer): 更新检查点保存和加载逻辑
|
2026-01-08 19:04:08 +08:00 |
|
test_dataset.py
|
fix(mmap): 修复样本数与键值计算逻辑并增强错误处理
|
2025-12-15 09:27:29 +08:00 |
|
test_early_stopping.py
|
fix(trainer): 更新检查点保存和加载逻辑
|
2026-01-08 19:04:08 +08:00 |
|
test_module.py
|
feat(model): 添加QK归一化和门控注意力支持
|
2026-01-05 16:14:44 +08:00 |
|
test_parallel.py
|
refactor(parallel): 重构parallel模块
|
2025-12-13 22:16:17 +08:00 |
|
test_sampler.py
|
fix(data/sampler): 修正拼写错误并增强采样器功能
|
2025-11-27 19:43:36 +08:00 |
|
test_tie_weight.py
|
feat(model): 添加QK归一化和门控注意力支持
|
2026-01-05 16:14:44 +08:00 |
|
test_train_config.py
|
fix(trainer): 修复参数传递问题和检查点保存问题
|
2025-12-08 13:28:11 +08:00 |
|
test_train_strategy.py
|
refactor(trainer): 优化trainer 结构
|
2025-12-07 21:23:05 +08:00 |
ViperEkura
d407962ffa