ByteDance-Seed / ByteDance-Seed/Bagel
如何获得推理时各个layer各个head的attention score？
Open
- Dominant language
- Python
- Stars
- 6.2k
- Forks
- 545
- PR merge metrics
- No merged PRs in 30d
Description
请问有直接调用的函数吗？开源代码封装太多了，需要修改的地方太多
Contributor guide
No contributing guide indexed for this repository
Research direction
Start by locating Bagel's inference entry point and the attention implementation, then trace whether layer- and head-level attention scores are retained or discarded. Confirm the available callable interface and the scope of any required change; done means the inference path exposes the requested scores or clearly documents that it does not.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- ai, machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100