feat: integrate the llama3 (8B, 70B) served by Groq #531

Appointat · 2024-04-24T18:33:12Z

Description

Describe your changes in detail.

Motivation and Context

Types of changes

What types of changes does your code introduce? Put an x in all the boxes that apply:

New feature (non-breaking change which adds core functionality)
Example (update in the folder of example)

coderabbitai · 2024-04-27T21:29:12Z

Important

Auto Review Skipped

Auto reviews are disabled on this repository.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

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>.
- 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 generate interesting stats about this repository and render them as a table.
- @coderabbitai show all the console.log statements in this repository.
- @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.

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 as PR comments)

@coderabbitai pause to pause the reviews on a PR.
@coderabbitai resume to resume the paused reviews.
@coderabbitai review to trigger a review. This is useful when automatic reviews are disabled for the repository.
@coderabbitai resolve resolve all the CodeRabbit review comments.
@coderabbitai help to get help.

Additionally, you can add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.

CodeRabbit Configration 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.

Appointat · 2024-04-28T17:23:30Z

@Wendong-Fan Hi, could you please help me add GROQ_API_KEY to the GitHub secret (I will DM you the secrete key value)? Thanks a lot in advance.

ocss884

LGTM! Left a question

ocss884 · 2024-04-30T07:15:33Z

camel/models/groq_model.py

+        if not self._token_counter:
+            # Groq API does not provide any token counter, so we use the
+            # OpenAI token counter as a placeholder.
+            self._token_counter = OpenAITokenCounter(ModelType.GPT_3_5_TURBO)


Why not use Autotokenizer to load the correct llama3 tokenizer?

Good idea, and I will try it.

Hi, I have implemented the Autotokenizer. wait for your review

Wendong-Fan

Thanks @Appointat 's contribution! Overall is great, left some comments

camel/models/__init__.py

Wendong-Fan · 2024-05-06T15:09:36Z

camel/utils/token_counting.py

+class GroqLlama3TokenCounter(BaseTokenCounter):
+    def __init__(self, model_type: ModelType):
+        r"""Constructor for the token counter for Llama3 models served by Groq.
+
+        Args:
+            model_type (ModelType): Model type for which tokens will be
+                counted.
+        """
+
+        self.model_type = model_type
+        # Since Groq API does not provide any token counter, we use the
+        # AutoTokenizer from transformers as a placeholder.
+        self.tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')
+
+    def count_tokens_from_messages(self, messages: List[OpenAIMessage]) -> int:
+        r"""Count number of tokens in the provided message list using
+        loaded tokenizer specific for this type of model.
+
+        Args:
+            messages (List[OpenAIMessage]): Message list with the chat history
+                in Groq llama3 API format.
+
+        Returns:
+            int: Number of tokens in the messages.
+        """
+        num_tokens = 0
+        for message in messages:
+            for _, value in message.items():


can we use class OpenSourceTokenCounter do this?

Wendong-Fan · 2024-05-06T15:22:48Z

camel/models/groq_model.py

+        additional arguments to check.
+        Raises:


Suggested change

additional arguments to check.

Raises:

additional arguments to check.

Raises:

Wendong-Fan · 2024-05-06T15:57:19Z

camel/models/groq_model.py

+        _choices: List[Choice] = []
+        for choice in response.choices:
+            choice.message = ChatCompletionMessage(
+                role=choice.message.role,  # type: ignore[arg-type]
+                content=choice.message.content,
+                tool_calls=choice.message.tool_calls,  # type: ignore[arg-type]
+            )  # type: ignore[assignment]
+            _choice = Choice(**choice.__dict__)
+            _choices.append(_choice)
+        response.choices = _choices  # type: ignore[assignment]
+
+        response.usage = CompletionUsage(**response.usage.__dict__)  # type: ignore[assignment]
+        return ChatCompletion(**response.__dict__)


I'm confused with why we use Choice here, the response is already ChatCompletion. could you explain a little bit? thanks!

ocss884

left some comments

ocss884 · 2024-05-07T01:44:56Z

camel/utils/token_counting.py

+        self.model_type = model_type
+        # Since Groq API does not provide any token counter, we use the
+        # AutoTokenizer from transformers as a placeholder.
+        self.tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')


I suggest not launching any tokenizer if allowed. User won't go through every line of code but just find there is a tokenizer and use it. They can not notice it is not the correct one until they notice some errors in token number and try to figure out what's the problem.

Or, we need to specify which open-source model is launched and try to load it via HF libraries.

Appointat added 5 commits April 24, 2024 19:59

feat: add llama3 provided by groq

85eaced

feat: add llama3 provided by groq

4fef335

Merge remote-tracking branch 'origin' into feature/groq-llama3

2d78038

refactor code to use api_key_required decorator consistently

2590f5c

update poetry toml

85a55f2

Appointat marked this pull request as draft April 24, 2024 18:34

Appointat changed the title ~~feat: integrate the llama3 (8B, 70B) served by LLama3~~ feat: integrate the llama3 (8B, 70B) served by Groq Apr 24, 2024

Appointat self-assigned this Apr 24, 2024

Appointat added the Model Related to backend models label Apr 24, 2024

camel-ai deleted a comment from coderabbitai bot Apr 24, 2024

Merge remote-tracking branch 'origin' into feature/groq-llama3

4f307ad

Wendong-Fan added this to the Sprint 2 milestone Apr 28, 2024

Wendong-Fan linked an issue Apr 28, 2024 that may be closed by this pull request

[Feature Request] Integration of Llama 3 from Groq cloud API #520

Open

2 tasks

Appointat added 2 commits April 28, 2024 18:03

feat: add model configs, fix decorator of step function

f2e57f0

Merge branch 'master' into feature/groq-llama3

cff5e04

Appointat marked this pull request as ready for review April 28, 2024 16:05

add GROQ_API_KEY to yml

6996309

Wendong-Fan requested a review from a team April 29, 2024 11:53

ocss884 reviewed Apr 30, 2024

View reviewed changes

Appointat requested a review from a team April 30, 2024 22:16

Appointat added 7 commits May 5, 2024 17:40

Merge branch 'master' into feature/groq-llama3

136d0b7

update poetry

5fb90a7

remove type ignore

fec77a6

reformated by ruff

7b0f0b0

use AutoTokenizer as tokennizer

793a688

fix: update token_counter import in test_groq_llama3_model.py

eca647a

Merge branch 'master' into feature/groq-llama3

431f0f0

Wendong-Fan reviewed May 6, 2024

View reviewed changes

ocss884 reviewed May 7, 2024

View reviewed changes

fix: remove GroqModel from camel/models/__init__.py

936d6e4

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat: integrate the llama3 (8B, 70B) served by Groq #531

feat: integrate the llama3 (8B, 70B) served by Groq #531

Appointat commented Apr 24, 2024 •

edited

coderabbitai bot commented Apr 27, 2024

Auto Review Skipped

Chat

CodeRabbit Commands (invoked as PR comments)

CodeRabbit Configration File (`.coderabbit.yaml`)

Documentation and Community

Appointat commented Apr 28, 2024

ocss884 left a comment

ocss884 Apr 30, 2024

Appointat Apr 30, 2024

Appointat May 5, 2024

Wendong-Fan left a comment

Wendong-Fan May 6, 2024

Wendong-Fan May 6, 2024

Wendong-Fan May 6, 2024

ocss884 left a comment

ocss884 May 7, 2024

feat: integrate the llama3 (8B, 70B) served by Groq #531

Are you sure you want to change the base?

feat: integrate the llama3 (8B, 70B) served by Groq #531

Conversation

Appointat commented Apr 24, 2024 • edited

Description

Motivation and Context

Types of changes

coderabbitai bot commented Apr 27, 2024

Auto Review Skipped

Chat

CodeRabbit Commands (invoked as PR comments)

CodeRabbit Configration File (.coderabbit.yaml)

Documentation and Community

Appointat commented Apr 28, 2024

ocss884 left a comment

Choose a reason for hiding this comment

ocss884 Apr 30, 2024

Choose a reason for hiding this comment

Appointat Apr 30, 2024

Choose a reason for hiding this comment

Appointat May 5, 2024

Choose a reason for hiding this comment

Wendong-Fan left a comment

Choose a reason for hiding this comment

Wendong-Fan May 6, 2024

Choose a reason for hiding this comment

Wendong-Fan May 6, 2024

Choose a reason for hiding this comment

Wendong-Fan May 6, 2024

Choose a reason for hiding this comment

ocss884 left a comment

Choose a reason for hiding this comment

ocss884 May 7, 2024

Choose a reason for hiding this comment

Appointat commented Apr 24, 2024 •

edited

CodeRabbit Configration File (`.coderabbit.yaml`)