Skip to content
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

feat: deprecate Qwen1.5 and gemma; and replace them with Qwen2.5 and gemma2 #23

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

vatsalkshah
Copy link
Contributor

@vatsalkshah vatsalkshah commented Oct 18, 2024

deprecate Qwen1.5 and gemma; and replace them with Qwen2.5 and gemma2

Summary by CodeRabbit

  • New Features

    • Updated model identifiers to reflect the latest versions for improved performance and accuracy.
    • Adjusted configuration parameters for various model sizes to enhance training efficiency.
  • Documentation

    • Revised README.md to include updated model descriptions and identifiers.
    • Updated training_args.yaml to align with the new model configurations and parameters.

Copy link

coderabbitai bot commented Oct 18, 2024

Walkthrough

The pull request includes updates to the README.md and training_args.yaml files, specifically altering model identifiers and descriptions for the submission of model results. The hg_repo_id and base_model fields in the JSON data structure have been changed to reflect a newer model version. Additionally, the training_args.yaml file has seen modifications in model names and parameters, including adjustments to lora_rank and lora_alpha for different model configurations.

Changes

File Change Summary
README.md Updated hg_repo_id from "Qwen/Qwen1.5-1.8B-Chat" to "Qwen/Qwen2.5-1.5B"; updated base_model from "qwen1.5" to "qwen2.5".
training_args.yaml Updated model identifiers from Qwen/Qwen1.5-* to Qwen/Qwen2.5-*; adjusted lora_rank and lora_alpha parameters; updated Google model identifiers from google/gemma-* to google/gemma-2-*.

Possibly related PRs

Poem

🐇 In the meadow, changes bloom,
New models rise, dispelling gloom.
Qwen hops to a brighter day,
With lora settings on display.
From one to two, the journey flows,
In the fields where knowledge grows! 🌼


📜 Recent review details

Configuration used: CodeRabbit UI
Review profile: CHILL

📥 Commits

Files that changed from the base of the PR and between 67a41e4 and 7438977.

📒 Files selected for processing (3)
  • README.md (1 hunks)
  • training_args.yaml (1 hunks)
  • utils/constants.py (1 hunks)
🔥 Files not summarized due to errors (1)
  • utils/constants.py: Error: Server error: no LLM provider could handle the message
🧰 Additional context used
🔇 Additional comments (7)
training_args.yaml (6)

1-7: LGTM! Model update and configuration look good.

The update from Qwen1.5-0.5B to Qwen2.5-0.5B aligns with the PR objectives. The configuration parameters are consistent with other models, except for lora_rank and lora_alpha, which are higher for this smaller model (8 and 16, respectively). This difference might be intentional to allow more flexibility in fine-tuning the smaller model.


17-23: LGTM! Model update and configuration are consistent.

The update from Qwen1.5-7B to Qwen2.5-7B aligns with the PR objectives. The configuration is identical to the 1.5B model, maintaining consistency across larger models. The lora_rank and lora_alpha values (4 and 8, respectively) are appropriate for this model size.


25-31: LGTM! Model update and configuration look good.

The update from google/gemma-2b to google/gemma-2-2b aligns with the PR objectives. The naming convention change (adding "-2-" in the model name) is consistent with the deprecation of the original Gemma model and the introduction of Gemma 2. The configuration parameters are identical to the Qwen 1.5B and 7B models, maintaining consistency across different model families.


Line range hint 33-39: LGTM! Model update and configuration are consistent.

The update from google/gemma-7b to google/gemma-2-7b aligns with the PR objectives. The naming convention change (adding "-2-" in the model name) is consistent with the gemma-2-2b model and reflects the introduction of Gemma 2. The configuration parameters are identical to the gemma-2-2b model and the larger Qwen models, maintaining consistency across different model families and sizes.


Line range hint 1-39: Summary: Changes align well with PR objectives and maintain consistency.

The updates in this file successfully implement the PR objectives:

  1. Qwen1.5 models have been replaced with Qwen2.5 models.
  2. Gemma models have been updated to Gemma 2 models.

The configurations maintain consistency across different model sizes and families, with minor variations for the smallest Qwen model (0.5B). This standardized approach to fine-tuning across models is commendable.

Overall, the changes are well-implemented and maintain the structure and formatting of the original file.


9-15: LGTM! Model update and configuration look good, with a note on model size change.

The update from Qwen1.5-1.8B to Qwen2.5-1.5B aligns with the PR objectives. Note that the model size has changed from 1.8B to 1.5B parameters. The lora_rank and lora_alpha values (4 and 8, respectively) are lower compared to the 0.5B model, which is consistent with the larger models in this configuration.

To ensure this change is reflected correctly throughout the codebase, please run the following script:

✅ Verification successful

Verified! No remaining references to the 1.8B model size found.

The model update from Qwen1.5-1.8B to Qwen2.5-1.5B is consistent throughout the codebase. All references to the previous model size have been successfully removed or updated.

🏁 Scripts executed

The following scripts were executed for the analysis:

Script:

#!/bin/bash
# Description: Verify the model size change from 1.8B to 1.5B is consistent across the codebase

# Test: Search for any remaining references to the 1.8B model size
rg --type yaml --type markdown --type python "Qwen.*1\.8B"

Length of output: 91


Script:

#!/bin/bash
# Description: Verify the model size change from 1.8B to 1.5B is consistent across the codebase

# Test: Search for any remaining references to the 1.8B model size without specifying file types
rg "Qwen.*1\.8B"

Length of output: 18

README.md (1)

78-79: ⚠️ Potential issue

Update model identifiers and fix JSON formatting

The changes correctly reflect the update from Qwen1.5 to Qwen2.5, aligning with the PR objectives. However, there's a minor formatting issue in the JSON structure.

Please apply the following changes:

-        "hg_repo_id": "Qwen/Qwen2.5-1.5B,
-        "base_model": "qwen2.5",
+        "hg_repo_id": "Qwen/Qwen2.5-1.5B",
+        "base_model": "qwen2.5",

This fix ensures proper JSON formatting by adding the missing closing quote for the hg_repo_id value.

To ensure consistency across the document, let's check for any other mentions of Qwen1.5 that might need updating:

✅ Verification successful

JSON formatting corrected and Qwen1.5 references updated

All instances of "Qwen1.5" have been successfully updated to "Qwen2.5," and the JSON formatting issue has been resolved.

🏁 Scripts executed

The following scripts were executed for the analysis:

Script:

#!/bin/bash
# Search for any remaining mentions of Qwen1.5 in the README.md file
rg "Qwen1\.5" README.md

Length of output: 25


Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media?

❤️ Share
🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

  • Review comments: Directly reply to a review comment made by CodeRabbit. Example:
    • I pushed a fix in commit <commit_id>, please review it.
    • Generate unit testing code for this file.
    • Open a follow-up GitHub issue for this discussion.
  • Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
    • @coderabbitai generate unit testing code for this file.
    • @coderabbitai modularize this function.
  • PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
    • @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
    • @coderabbitai read src/utils.ts and generate unit testing code.
    • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
    • @coderabbitai help me debug CodeRabbit configuration file.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

  • @coderabbitai pause to pause the reviews on a PR.
  • @coderabbitai resume to resume the paused reviews.
  • @coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
  • @coderabbitai full review to do a full review from scratch and review all the files again.
  • @coderabbitai summary to regenerate the summary of the PR.
  • @coderabbitai resolve resolve all the CodeRabbit review comments.
  • @coderabbitai configuration to show the current CodeRabbit configuration for the repository.
  • @coderabbitai help to get help.

Other keywords and placeholders

  • Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
  • Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
  • Add @coderabbitai anywhere in the PR title to generate the title automatically.

CodeRabbit Configuration File (.coderabbit.yaml)

  • You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
  • Please see the configuration documentation for more information.
  • If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Documentation and Community

  • Visit our Documentation for detailed information on how to use CodeRabbit.
  • Join our Discord Community to get help, request features, and share feedback.
  • Follow us on X/Twitter for updates and announcements.

"hg_repo_id": "Qwen/Qwen1.5-1.8B-Chat",
"base_model": "qwen1.5",
"hg_repo_id": "Qwen/Qwen2.5-1.5B,
"base_model": "qwen2.5",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

for base_model we are still using qwen1.5 since they stay the same

"Qwen/Qwen1.5-7B": "qwen1.5",
"google/gemma-2b": "gemma",
"google/gemma-7b": "gemma",
"Qwen/Qwen2.5-0.5B": "qwen2.5",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same here

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants