mirror of
https://github.com/deepseek-ai/DeepSeek-Prover-V1.5.git
synced 2026-11-05 00:57:20 +00:00
[GH-ISSUE #19] DeepSeek-Prover-V1.5-Base performs significantly worse than DeepSeek-Prover-V1 in Table 1 #12
Labels
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
deepseek-ai/DeepSeek-Prover-V1.5#12
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @vadimkantorov on GitHub (Apr 10, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-Prover-V1.5/issues/19
Hi! Thanks for open-sourcing!
What is the intuition in Tab. 1 /
Single-Pass Whole-Proof Generation MethodsforDeepSeek-Prover-V1.5-Baseto fare much worse thanDeepSeek-Prover-V1: (@ 128): 29.7% vs 46.1%If I understood well from Sec. 2.1,
DeepSeek-Prover-V1.5-Basewas obtained by fine-tuning DeepSeek-Prover-V1. Is it correct? (or was it trained from scratch? or initialized from some other pretrained model? maybeDeepSeekMath 7B?) If so, what is the explanation for such degradation?Thank you!