Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Always fails to execute the action, and it responds with 'nothing happens'. #41

Open
prettystar0203 opened this issue Apr 15, 2024 · 4 comments

Comments

@prettystar0203
Copy link

For example, like this.This is from my trial_0.log:

go to cabinet 3
On the cabinet 3, you see a cup 1.
think: I found a cup in cabinet 3. I can use it to clean the plate. Next, I need to put the plate in countertop 1.
OK.
【> take cup 1 from cabinet 3
Nothing happens.】
It seems there is no cup in cabinet 3. Let's try another way to clean the plate.
Nothing happens.
go to cabinet 4
On the cabinet 4, you see nothing.

go to drawer 2
Nothing happens.
go to drawer 2
On the drawer 2, you see a pencil 1.
think: I need to keep looking for the bowl. Let's try drawer 3.
Nothing happens.
【go to drawer 3
Nothing happens.
go to drawer 3
Nothing happens.】

I would like to know if I am doing something wrong, and I would be very grateful if you could help me solve this issue.

@TuZzzz1
Copy link

TuZzzz1 commented Jun 3, 2024

I guess maybe the LLM is simulating the conversation itself.The content before "Nothing happens" are all produced by the LLM, but are not the interaction between LLM and the environment.
Like:go to cabinet 3
On the cabinet 3, you see a cup 1.
think: I found a cup in cabinet 3. I can use it to clean the plate. Next, I need to put the plate in countertop 1.
OK.
【> take cup 1 from cabinet 3
Contents above are actually produced by LLM, but it looks like it is the interaction between LLM and alfworld.

@Zxy-MLlab
Copy link

No, "No happens" is indeed the result of the environment return, I also encountered this problem, I wonder if you have solved it?

@lollipopwsy
Copy link

@TuZzzz1 @prettystar0203 @Zxy-MLlab Hello, I successfully got it to display
' Initializing AlfredTWEnv...
100%|| 341/341 [00:00<00:00, 947.27it/s]
Overall we have 134 games in split=eval_out_of_distribution
Evaluating with 134 games',

but it always displays 'Nothing happens'.

Details are as follows: You are in the middle of a room. Looking quickly around you, you see a cabinet 6, a cabinet 5, a cabinet 4, a cabinet 3, a cabinet 2, a cabinet 1, a coffeemachine 1, a countertop 3, a countertop 2, a countertop 1, a drawer 3, a drawer 2, a drawer 1, a fridge 1, a garbagecan 1, a microwave 1, a shelf 3, a shelf 2, a shelf 1, a sinkbasin 1, a stoveburner 4, a stoveburner 3, a stoveburner 2, a stoveburner 1, and a toaster 1.
Your task is to: clean some bowl and put it in cabinet.
Sending prompt to model...
Received response: > think: To solve the task, I need to find a bowl, clean it with the sinkbasin, and then put it in one of the cabinets. think: To solve the task, I need to find a bowl, clean it with the sinkbasin, and then put it in one of the cabinets.
Nothing happens.
Sending prompt to model...
Response too short, retrying (attempt 1)...
Received response: OK. Let's start by finding a bowl. Bowls are likely to be found in the cabinets, on the countertops, or in the drawers. I will check the cabinets first, starting with cabinet 1.
OK. Let's start by finding a bowl. Bowls are likely to be found in the cabinets, on the countertops, or in the drawers. I will check the cabinets first, starting with cabinet 1.
Nothing happens.
Sending prompt to model...
Received response: > go to cabinet 1
go to cabinet 1
Nothing happens.
Sending prompt to model...
Received response: Let's continue with our search for a bowl. I'll check the next cabinet.
Let's continue with our search for a bowl. I'll check the next cabinet.
Nothing happens.
Sending prompt to model...
Received response: > go to cabinet 2
go to cabinet 2
Nothing happens.
Sending prompt to model...
Received response: > go to cabinet 3
go to cabinet 3
Nothing happens.
Sending prompt to model...
Received response: > go to cabinet 4
go to cabinet 4
Nothing happens.
Sending prompt to model...
Received response: > go to cabinet 5
go to cabinet 5
Nothing happens.

Could you help me solve this? Thank you very much!

@TuZzzz1
Copy link

TuZzzz1 commented Feb 19, 2025

No, "No happens" is indeed the result of the environment return, I also encountered this problem, I wonder if you have solved it?

@TuZzzz1 @prettystar0203 @Zxy-MLlab Hello, I successfully got it to display ' Initializing AlfredTWEnv... 100%|| 341/341 [00:00<00:00, 947.27it/s] Overall we have 134 games in split=eval_out_of_distribution Evaluating with 134 games',

but it always displays 'Nothing happens'.

Details are as follows: You are in the middle of a room. Looking quickly around you, you see a cabinet 6, a cabinet 5, a cabinet 4, a cabinet 3, a cabinet 2, a cabinet 1, a coffeemachine 1, a countertop 3, a countertop 2, a countertop 1, a drawer 3, a drawer 2, a drawer 1, a fridge 1, a garbagecan 1, a microwave 1, a shelf 3, a shelf 2, a shelf 1, a sinkbasin 1, a stoveburner 4, a stoveburner 3, a stoveburner 2, a stoveburner 1, and a toaster 1. Your task is to: clean some bowl and put it in cabinet. Sending prompt to model... Received response: > think: To solve the task, I need to find a bowl, clean it with the sinkbasin, and then put it in one of the cabinets. think: To solve the task, I need to find a bowl, clean it with the sinkbasin, and then put it in one of the cabinets. Nothing happens. Sending prompt to model... Response too short, retrying (attempt 1)... Received response: OK. Let's start by finding a bowl. Bowls are likely to be found in the cabinets, on the countertops, or in the drawers. I will check the cabinets first, starting with cabinet 1. OK. Let's start by finding a bowl. Bowls are likely to be found in the cabinets, on the countertops, or in the drawers. I will check the cabinets first, starting with cabinet 1. Nothing happens. Sending prompt to model... Received response: > go to cabinet 1 go to cabinet 1 Nothing happens. Sending prompt to model... Received response: Let's continue with our search for a bowl. I'll check the next cabinet. Let's continue with our search for a bowl. I'll check the next cabinet. Nothing happens. Sending prompt to model... Received response: > go to cabinet 2 go to cabinet 2 Nothing happens. Sending prompt to model... Received response: > go to cabinet 3 go to cabinet 3 Nothing happens. Sending prompt to model... Received response: > go to cabinet 4 go to cabinet 4 Nothing happens. Sending prompt to model... Received response: > go to cabinet 5 go to cabinet 5 Nothing happens.

Could you help me solve this? Thank you very much!
It's been a long time so I may forget how to solve the problem. But you can try check the output of LLM to see if it is in the right format as we wish. The environment return "No happens" means the response of LLM is not in the right format as the code request. You guys can try change to another llm like online stronger ones.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants