Efficient Multimodal Transfer of Reasoning-Capable LLMs
Nobody has claimed this yet.
Assessment
- Difficulty
- 5/5
- Estimated time
- Over a week
- Newbie friendliness
- 20/100
- Issue type
- Documentation
- Clarity
- Needs clarification
- Activity status
- Stale
- Domain
- ai, machine-learning
Research direction
No files, entry points, or tests are identified. First clarify whether the issue seeks experimental results, a theoretical explanation, or documentation about directly using f_l after SFT instead of initializing with f_l^s; done would require a documented comparison and explanation supported by the project’s methodology or experiments.
Written by the indexing model from the issue text.
Description
如果不用无推理能力的llm $$f_{l}^{s}$$ 先进行initial，而直接使用$$f_{l}$$sft以及后续过程的结果是怎样的呢？以及为什么这样直接迁移会取得比较好的性能，请问如何来解释这个现象呢？
- Dominant language
- Python
- Stars
- 3.2k
- Forks
- 282
- PR merge metrics
- No merged PRs in 30d
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from SkyworkAI/Skywork-R1V
-
Difficulty 5/5 Over a week Newbie friendliness 20/100
SkyworkAI/Skywork-R1V#61 ·
-
Difficulty 5/5 Over a week Newbie friendliness 20/100
SkyworkAI/Skywork-R1V#60 ·
-
Difficulty 5/5 Over a week Newbie friendliness 20/100
SkyworkAI/Skywork-R1V#59 ·
-
Difficulty 5/5 Over a week Newbie friendliness 25/100
SkyworkAI/Skywork-R1V#58 ·
-
Difficulty 5/5 Over a week Newbie friendliness 20/100
SkyworkAI/Skywork-R1V#57 ·
All issues in SkyworkAI/Skywork-R1V
Similar issues
-
sponsored
Difficulty 2/5 1-3 hours Newbie friendliness 65/100
-
Difficulty 2/5 1-3 hours Newbie friendliness 86/100
Diaoul/subliminal#1382 ·
-
Difficulty 1/5 Under an hour Newbie friendliness 92/100
-
triage/confirmed
Difficulty 2/5 1-3 hours Newbie friendliness 88/100
agentscope-ai/agentscope#2775 ·
-
Difficulty 2/5 1-3 hours Newbie friendliness 84/100