Add model_prompt config param for LLMBlock
#141
Merged
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
mergify
bot
added
documentation
|
I've added commit 2ff97ba to allow |
markmc
approved these changes
Jul 15, 2024
Our intiial pipeline configuration is version `1.0`. We haven't actually released anything yet, but I think it would be good to go ahead and get in the habit of managing this verison number as we make additions. Create a doc where we can keep a list of what changed at each version bump. Signed-off-by: Russell Bryant <[email protected]>
We previously always set the model prompt based on the model family. We have some cases where we'd like to override this default behavior explicitly as part of pipeline configuration. This parameter will do that. Signed-off-by: Russell Bryant <[email protected]>
Add a test case that ensures that passing `model_prompt` as an empty string results in `self.model_prompt` as empty. If this parameter is not specified (is None), `self.model_prompt` should be a non-empty string as a result of our default behavior of picking a prompt based on the model family. Signed-off-by: Russell Bryant <[email protected]>
An anticipated use case for model_prompt is simply to disable
this additional prompt. Currently the pipeline author would
need to specify model_prompt="{prompt}" to achieve this.
Make this easier by allowing model_prompt="" to have this
meaning
Signed-off-by: Mark McLoughlin <[email protected]>
|
|
||
| ## Pipeline Configuration Schema | ||
|
|
||
| A schema for validating pipeline configuration can be found in [`src/instructlab/sdg/pipelines/schema/v1.json`](../src//instructlab/sdg/pipelines/schema/v1.json) |
There was a problem hiding this comment.
nit:
double "/"'s in path name
jwm4
pushed a commit
to jwm4/sdg
that referenced
this pull request
Dec 13, 2024
…sign-doc Add design doc for training to explain accelerate inclusion for FSDP
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.
Closes #136
2363dfe docs: Add a doc to track pipeline config version history
87871d3 llmblock: add
model_promptconfig parameter4f042ae Add test case for LLMBlock model_prompt parameter
7045456 llmblock: allow model_prompt=""
commit 2363dfe
Author: Russell Bryant [email protected]
Date: Mon Jul 15 14:35:26 2024 -0400
commit 87871d3
Author: Russell Bryant [email protected]
Date: Mon Jul 15 14:40:18 2024 -0400
commit 4f042ae
Author: Russell Bryant [email protected]
Date: Mon Jul 15 14:57:32 2024 -0400
commit 7045456
Author: Mark McLoughlin [email protected]
Date: Mon Jul 15 23:56:10 2024 +0100