Add fine-tuning code and scripts #1
Merged
Conversation
This file contains hidden or 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
llewelld
force-pushed
the
baskerville
branch
5 times, most recently
from
5779493 to
f9c710c
Compare
Adds code for fine-tuning the models on Baskerville.
The directory contains both prediction and fine-tuning, so the name no longer makes sense. Keeping both tasks in the same directory turned out to be convenient because the downloads directory can be reused more easily by both in this case.
Wraps the model in FSDP, but still runs out of memory when performing the backwards step.
Adds headers to various files: 1. Shebang interpreter directive. 2. vim modeline configuration. 3. SPDX licence identifier.
Renames the scripts that match closest to those working on DAWN to use the suffix "aligned" to avoid confusion over thier purpose.
Adds a loss function to align with the formula in the paper.
Fixes the loss function implementation for the aligned and FSDP cases.
Adds code for comparing DAWN and Baskerville results and generating comparisong graphs.
Adds a README with instructions for how to run the comparisons and generate the comparison graphs.
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Outdated suggestions cannot be applied.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.
Adds the initial code for fine-tuning.
Only the small model works on an 80 GiB A100. The standard model gives an out-of-memory error on the GPU.