add max_tokens field for llm interface

[yisz]

🚀 This description was created by Ellipsis for commit a9e8583

Summary:

This PR introduces a max_tokens parameter to the LLM interface methods, allowing control over the number of tokens generated, and adds Cohere API key support in the environment configuration.

Key points:

• Added max_tokens parameter to LLMInterface.run and LLMFactory.run with a default of 1024.
• Updated _llm_response in LLMFactory to pass max_tokens to LLM client calls.
• Added COHERE_API_KEY to .env.example.

---

Generated with ❤️ by ellipsis.dev

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to a9e8583 in 33 seconds

More details

• Looked at 112 lines of code in 2 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. continuous_eval/llm_factory.py:42

• Draft comment:
  It's important to include a description in the PR to explain the reason for these changes and how they improve the functionality or fix an issue. This helps reviewers understand the context and the intent behind the changes.
• Reason this comment was not posted:
  Confidence changes required: 33%
  The PR description is empty, which is not ideal because it doesn't provide context or reasoning behind the changes. However, the code changes suggest that the max_tokens parameter is being added to the run method of the LLMInterface and its implementations. This parameter controls the maximum number of tokens the model should generate in response to a prompt. The changes are consistent across the file, and the default value for max_tokens is set to 1024, which is a reasonable default for many language models.

Workflow ID: wflow_NABLkaixgi063xd4

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.