-
-
Notifications
You must be signed in to change notification settings - Fork 4
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
Optimise the output to reduce token consumption #113
Open
mihaistate
wants to merge
6
commits into
esbmc:master
Choose a base branch
from
mihaistate:master
base: master
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Commits on Feb 17, 2024
-
Created a new method for output optimisation
Implemented a function in esbmc_util that uses regex patterns to remove some output lines to reduce token consmuption and to facilitate the reduction of the likelihood that LLMs like GPT 3.5 turbo reach the token limit for some codebases that need comprehensive explanations. Addded in config.json four new cases that address common bugs in C programs: buffer overflow, arithmetic overflow, array out-of-bounds and memory leaks. Signed-off-by: mihai.state <mihaita.state@yahoo.com>
Configuration menu - View commit details
-
Copy full SHA for 783f6b1 - Browse repository at this point
Copy the full SHA 783f6b1View commit details
Commits on Feb 18, 2024
-
Configuration menu - View commit details
-
Copy full SHA for d8f6704 - Browse repository at this point
Copy the full SHA d8f6704View commit details
Commits on Mar 5, 2024
-
Improved the ESBMC-AI output reduction
Created more regex patterns that remove some lines that can be ommitted and provide shorter outputs. Added a few more indications in the config.json file that tell the LLM to avoid mentioning time measurements Signed-off-by: mihai.state <mihaita.state@yahoo.com>
Configuration menu - View commit details
-
Copy full SHA for 1dc5c0b - Browse repository at this point
Copy the full SHA 1dc5c0bView commit details
Commits on Mar 25, 2024
-
Created new output scenarious in the config.json file for NaN check, struct fields check, deadlock check and data races check Created the reduce_output field in the config.json file to tell the LLM to produce a smaller output Created new regex patterns in the reduce_output2() function Created the remove_patterns_nltk that uses NLP methods to identify patterns in a sequence of tokenized words. The advantage of it are that it provides more granular control over text sentences, it can be more precise in some situations. Created the test_output_reducer file that tests the functionality of the esbmc_output_optimisation(), reduce_output2() and remove_patterns_nltk() functions that covers all the string patterns Instantiated the GPT_4_TURBO_PREVIEW model in the enum of the ai_models.py and used it for various code bases. Signed-off-by: mihai.state <mihaita.state@yahoo.com>
Configuration menu - View commit details
-
Copy full SHA for abbf9a0 - Browse repository at this point
Copy the full SHA abbf9a0View commit details
Commits on Mar 26, 2024
-
Configuration menu - View commit details
-
Copy full SHA for c7ded6f - Browse repository at this point
Copy the full SHA c7ded6fView commit details
Commits on Jun 27, 2024
-
Optimized the regex pattens to be more time efficient
Removed the 'flags=re.MULTILINE' argument for the regex patterns that work properly without it Used the re.compile method for some patterns to reduce time complexity Signed-off-by: mihai.state <mihaita.state@yahoo.com>
Configuration menu - View commit details
-
Copy full SHA for 36798da - Browse repository at this point
Copy the full SHA 36798daView commit details
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.