karthikv792 / llms-planning Goto Github PK
View Code? Open in Web Editor NEWAn extensible benchmark for evaluating large language models on planning
An extensible benchmark for evaluating large language models on planning
Hi,
May I ask what's the Linux system you are using? I have been trying to compile the pr2plan package on my Ubuntu 18.04 & 20.4 systems, but I keep getting errors about lib files missing.
Thanks.
I tried to run the provided example in /gpt_plan_test
python3 ReasoningTasks.py --task t1 --engine davinci --verbose True
in Google Colab, but am receiving the error:
File "/content/gpt-plan-benchmark/gpt_plan_test/utils/__init__.py", line 294, in get_plan_as_text
with open(plan_file) as f:
FileNotFoundError: [Errno 2] No such file or directory: 'sas_plan
I can of course add a file "sas_plan" myself manually, but I have no idea what is supposed to be in it since the README does not mention it. Is it supposed to be auto-generated?
I run the code by "python3 llm_plan_pipeline.py --task 't1' --config 'blocksworld' --engine ada", but get error as
Traceback (most recent call last):
File "/home/hchen/LLMs_planning/plan-bench/llm_plan_pipeline.py", line 66, in
prompt_generator.task_1_plan_generation(specified_instances, random_example)
File "/home/hchen/LLMs_planning/plan-bench/prompt_generation.py", line 152, in task_1_plan_generation
gt_plan_text = get_plan_as_text(self.data)
File "/home/hchen/LLMs_planning/plan-bench/utils/pddl_to_text.py", line 131, in get_plan_as_text
with open(plan_file) as f:
FileNotFoundError: [Errno 2] No such file or directory: 'sas_plan'
does anyone have the same bug? and how do you fix this?
Appreciate the repo and contribution. I met the following issue when trying to generate the prompts.
there is no sas_plan file. There were two issues on this bug and I checked them carefully, but I didn't find a solution from there.
If I manually create a sas_plan file and copy the lines from gpt_sas_file and paste them in sas_plab. All the ground truths in the prompts file become the same. Like this "ground_truth_plan": "(unstack yellow orange)\n(put-down yellow)\n(pick-up orange)\n(stack orange red)\n".
I guess the sas_plan is for generating ground truth in pddl. Please help me and I really appreciate your effort. Thank you.
How should I set the path of FAST_DOWNWARD? The link (https://www.fast-downward.org/ObtainingAndRunningFastDownward) is empty.
The Requirements in the README of plan-bench about Fast Downward, VAL and PR2Plan related links jump to empty pages and cannot be downloaded.
In addition, the CONFIG.Yaml file does not exist in the /plan-bench/config/ file
I think there are a few bugs in the problem_generator.py.
For example,
self.data is� None (
Could you please help to fix the bugs in problem_generator.py?
I have the Wikihow latest dataset from Kaggle in csv file format also I have one medical data. I want to create similar Pkl file that you created. Could you please share the code that generate these Pkl file.
I am trying to evaluate the responses.
I downloaded the repo and set os.environ['VAL'] = '/home/wenjun/2024PlanBench/planner_tools/VAL. However, when I tried to evaluate the responses, I got the error "/bin/sh: 1: /home/wenjun/2024PlanBench/planner_tools/VAL/validate: Permission denied".
Should I install the validate package or something?
Please help me and thank you in advance.
While running gpt-actions.py I found out if block if cnt > 0
from line 109 to 133 is not executing because of line 80 cnt -= 1
.
Is it intentional or some logic behind it.
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.