wellecks / llemma_formal2formal Goto Github PK
View Code? Open in Web Editor NEWLlemma formal2formal (tactic prediction) theorem proving experiments
License: MIT License
Llemma formal2formal (tactic prediction) theorem proving experiments
License: MIT License
I wanted code like this one https://github.com/wellecks/llemma_formal2formal/blob/f96fdb1642b2f21bed85c19640b6f1511abb1116/proofsearch.py but for other proof search algs, any exist?
Hi, I followed the instruction in readme to download nightly version but find a version error
root@c03bdf811de6:/code/llemma_formal2formal# python
Python 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from lean_dojo import LeanGitRepo
2024-08-03 10:28:54.469 | WARNING | lean_dojo:<module>:33 - Running LeanDojo as the root user may cause unexpected issues. Proceed with caution.
>>> repo = LeanGitRepo('https://github.com/rah4927/lean-dojo-mew', "d00c776260c77de7e70125ef0cd119de6c0ff1de")
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "<string>", line 5, in __init__
File "/usr/local/lib/python3.10/dist-packages/lean_dojo/data_extraction/lean.py", line 434, in __post_init__
lean_version = get_lean4_commit_from_config(config)
File "/usr/local/lib/python3.10/dist-packages/lean_dojo/data_extraction/lean.py", line 336, in get_lean4_commit_from_config
return _to_commit_hash(LEAN4_REPO, version)
File "/usr/local/lib/python3.10/dist-packages/lean_dojo/data_extraction/lean.py", line 100, in _to_commit_hash
raise ValueError(f"Invalid tag or branch: `{label}` for {repo}")
ValueError: Invalid tag or branch: `nightly-2023-08-19` for Repository(full_name="leanprover/lean4")
Do you know the reason? Should I download another version of elan?
Thank you for your valuable contribution to formal theorem proving. I would like to cite your work but I didn't find a reported result on the pass rate of minif2f-valid. I am sorry I don't have enough resources now to reproduce the results. So I wonder if you have done any evaluation on minif2f-valid. What is the performance? Thank you in advance!
A better architecture might be that the language model can determine possible methods to solve mathematical problems, then perform specific operations to transform the solution of the mathematical problem into a series of thoughts, fill the thought chain, and add generality in the transformation part to make the thinking more divergent, and then the part of filling the thought chain needs to be more accurate
It's a bit like the architecture of chatgpt+mathematica, but the problem with the chatgpt+mathematica architecture is that mathematica is too biased towards hard coding and often reports errors, and chatgpt does not specifically train to split the input of a mathematical problem into step-by-step solutions
Is that all greedy decoding?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.