update eval #450
Merged
update eval #450
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: n1ck-guo <[email protected]>
There was a problem hiding this comment.
PR Overview
This PR updates the evaluation-related functionality and argument parsing in the project. Key changes include:
- Modifying argument aliases and default values for tasks and evaluation batch size.
- Refactoring and enhancing exception handling in evaluation functions to better manage out-of-memory errors.
- Minor adjustments in utility functions and autoround module to improve error detection.
Reviewed Changes
| File | Description |
|---|---|
| auto_round/script/llm.py | Updated argument names/aliases and refactored evaluation calls to use eval_sequence consistently. |
| auto_round/utils.py | Removed an extra blank line; no functional changes. |
| auto_round/main.py | Adjusted evaluation call parameters per updated interface. |
| auto_round/autoround.py | Enhanced error handling in caching intermediate data by extending memory error check. |
Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.
Comments suppressed due to low confidence (2)
auto_round/script/llm.py:201
- The introduction of both '--tasks' and '--task' may lead to confusion regarding which alias to use, especially when the default value implies a list. Consider standardizing on a single, consistent alias (e.g., '--tasks') throughout the code.
self.add_argument(
"--tasks",
"--task",
auto_round/script/llm.py:209
- Renaming the evaluation batch size argument from '--eval_bs' to '--batch_size' might introduce ambiguity with training batch size parameters defined elsewhere. Consider using distinct names for evaluation and training batch sizes to avoid potential misconfiguration.
self.add_argument("--batch_size", "--eval_bs", "--bs", default=8, type=int, help="batch size in evaluation")
wenhuach21
reviewed
Mar 3, 2025
wenhuach21
reviewed
Mar 3, 2025
wenhuach21
reviewed
Mar 3, 2025
| batch_size = 8 | ||
| if not isinstance(model, str): | ||
| parallelism = False | ||
| hflm = HFLM( |
There was a problem hiding this comment.
why not just call simple_eval_with_user_model
There was a problem hiding this comment.
trying use this func to replace the whole eval folder
Signed-off-by: n1ck-guo <[email protected]>
Signed-off-by: n1ck-guo <[email protected]>
Signed-off-by: n1ck-guo <[email protected]>
wenhuach21
reviewed
Mar 4, 2025
auto_round/script/llm.py
Outdated
| "--native_eval", | ||
| "--native", | ||
| action="store_true", | ||
| help="use the native lm_eval instead of eval task by task.") |
There was a problem hiding this comment.
use the native as default.
Signed-off-by: n1ck-guo <[email protected]>
wenhuach21
reviewed
Mar 6, 2025
auto_round/script/llm.py
Outdated
| self.add_argument( | ||
| "--disable_trust_remote_code", action='store_true', help="whether to disable trust_remote_code") | ||
| self.add_argument("--eval_bs", "--bs", "--batch_size", default=None, type=int, help="batch size in evaluation") | ||
| self.add_argument("--eval_bs", "--bs", "--batch_size", default=8, type=int, help="batch size in evaluation") |
There was a problem hiding this comment.
why change the default value
There was a problem hiding this comment.
default set to None, and if bs is None, will use auto.
wenhuach21
reviewed
Mar 6, 2025
auto_round/script/llm.py
Outdated
| @@ -580,14 +587,23 @@ def tune(args): | |||
|
|
|||
| if args.eval_bs is None or args.eval_bs == "auto": | |||
| args.eval_bs = 16 | |||
There was a problem hiding this comment.
could we set it to auto bs for task by task scenario
Signed-off-by: n1ck-guo <[email protected]>
wenhuach21
reviewed
Mar 6, 2025
| if args.eval_task_by_task: | ||
| eval_task_by_task(user_model, device=device_str, tasks=args.tasks, batch_size=args.eval_bs) | ||
| else: | ||
| if args.eval_bs is None: |
There was a problem hiding this comment.
better double check if it works well. I remember it has some issues.
There was a problem hiding this comment.
checked. Work well.
wenhuach21
approved these changes
Mar 6, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.