add tests for merging lora and validating the dtype #1512
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
NanoCode012
reviewed
Apr 12, 2024
Comment on lines
+107
to
+112
| cfg.lora_model_dir = cfg.output_dir | ||
| cfg.load_in_4bit = False | ||
| cfg.load_in_8bit = False | ||
| cfg.flash_attention = False | ||
| cfg.deepspeed = None | ||
| cfg.fsdp = None |
There was a problem hiding this comment.
This can be excluded as the modify_cfg_for_merge should've set it?
Suggested change
| cfg.lora_model_dir = cfg.output_dir | |
| cfg.load_in_4bit = False | |
| cfg.load_in_8bit = False | |
| cfg.flash_attention = False | |
| cfg.deepspeed = None | |
| cfg.fsdp = None |
| cfg.fsdp = None | ||
|
|
||
| cfg = modify_cfg_for_merge(cfg) | ||
| cfg.merge_lora = True |
There was a problem hiding this comment.
Let's move this setting inside the modify_cfg function as well.
| @@ -27,21 +28,26 @@ def do_cli(config: Path = Path("examples/"), **kwargs): | |||
| flash_attention=False, | |||
There was a problem hiding this comment.
If the above section already sets these properties, is it necessary to set it again below?
NanoCode012
reviewed
Apr 12, 2024
| # pylint: disable=duplicate-code | ||
| cfg = DictDefault( | ||
| { | ||
| "base_model": "JackFram/llama-68m", |
There was a problem hiding this comment.
Also, sometimes, this issue can occur for different model types. For ex, previous llama merge was fine, but mistral was not. Do we need to test this for other arch?
| cli_args = TrainerCliArgs() | ||
| dataset_meta = load_datasets(cfg=cfg, cli_args=cli_args) | ||
|
|
||
| train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta) |
There was a problem hiding this comment.
I don't think you need to train a model, maybe a tiny adapter can be uploaded to HF which we use for merge?
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.