Repository navigation
Fix Gemini Configuration Documentation for wren-ai-service - #1263
sebastien-boulle wants to merge 2 commits into
Conversation
WalkthroughThe configuration file for Google AI Studio has been updated. Both the LLM and embedder sections now include an explicit Changes
Suggested reviewers
Poem
Tip 🌐 Web search-backed reviews and chat
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? 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
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)
Other keywords and placeholders
CodeRabbit Configuration File (
|
…h Google documentation Based on https://ai.google.dev/gemini-api/docs/embeddings\?hl\=fr\#curl the model is not prefixed, and the url doesn't contain openai. Note also that the previsous version of the doc was not working because the model was set to text-embedding-004 in the pipeline which didn't match the previous label "gemini/text-embedding-004"
Without it the ai-service is not able to connect even if the key is in the .env
3a681f6 to
83e6746
Compare
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
wren-ai-service/docs/config_examples/config.google_ai_studio.yaml(2 hunks)
🧰 Additional context used
🪛 YAMLlint (1.35.1)
wren-ai-service/docs/config_examples/config.google_ai_studio.yaml
[warning] 20-20: wrong indentation: expected 0 but found 2
(indentation)
🔇 Additional comments (1)
wren-ai-service/docs/config_examples/config.google_ai_studio.yaml (1)
10-10: LLM Configuration – API Key Addition
The new keyapi_key_name: GEMINI_API_KEYfor the LLM model clearly documents which environment variable should be used for authentication. This improves the clarity of the configuration and aligns with the intended usage.
| # put GEMINI_API_KEY=<your_api_key> in ~/.wrenai/.env | ||
| - model: gemini/text-embedding-004 # gemini/<gemini_model_name> | ||
| api_base: https://generativelanguage.googleapis.com/v1beta/openai # change this according to your embedding model | ||
| - model: text-embedding-004 | ||
| api_base: https://generativelanguage.googleapis.com/v1beta/models | ||
| api_key_name: GEMINI_API_KEY | ||
| timeout: 120 |
There was a problem hiding this comment.
💡 Verification agent
🧩 Analysis chain
Embedder Configuration – Model Update & YAML Indentation Check
The embedder section now uses the simplified model name text-embedding-004, updates the api_base to https://generativelanguage.googleapis.com/v1beta/models, and adds the api_key_name: GEMINI_API_KEY, which are all positive improvements for clarity and consistency with Google’s Gemini API documentation.
However, YAMLlint reported an indentation warning on line 20 (“wrong indentation: expected 0 but found 2”). This suggests that the list item under the models: key may be indented inconsistently relative to the LLM section. Typically, list items should be indented uniformly. For example, if the LLM section’s list item starts flush (or with a defined two-space indent), then the embedder section should follow the same pattern. Additionally, the timeout: 120 on line 23 appears to be indented similarly to the list items—if it is supposed to be a top-level key for the embedder configuration, it may need to be outdented.
Please verify and adjust the embedder block’s indentation for consistency. Here’s a suggested diff if you intend to mimic the LLM section’s style:
18 models:
19 # put GEMINI_API_KEY=<your_api_key> in ~/.wrenai/.env
-23 timeout: 120
+20 - model: text-embedding-004
+21 api_base: https://generativelanguage.googleapis.com/v1beta/models
+22 api_key_name: GEMINI_API_KEY
+23 timeout: 120If the intended design is different (for example, keeping timeout within the embedder block), please ensure that the indentation conforms to YAML best practices and that YAMLlint passes without warnings.
Error: No shell scripts or web queries found
🏁 Script executed:
#!/bin/bash
# Display lines 15 to 30 from the YAML configuration file to inspect indentation.
sed -n '15,30p' wren-ai-service/docs/config_examples/config.google_ai_studio.yamlLength of output: 412
Embedder YAML Indentation Issue – Adjust the Placement of timeout
The configuration file shows that the timeout: 120 line is currently indented as part of the model definition in the models: list. However, YAMLlint reported an indentation warning (“wrong indentation: expected 0 but found 2”), which indicates that the timeout setting is expected to be a top-level key within the embedder block rather than nested inside the list item.
- If the intention is to apply
timeoutto the embedder configuration as a whole (mimicking the LLM section’s style), then it should be outdented so that it sits at the same level as themodels:key. - Update the embedder section similar to the suggested diff: move
timeout: 120to be flush withmodels:rather than being indented under the list item.
Please review whether timeout should be a property for the individual model or for the embedder configuration overall and adjust the YAML accordingly to satisfy YAMLlint.
🧰 Tools
🪛 YAMLlint (1.35.1)
[warning] 20-20: wrong indentation: expected 0 but found 2
(indentation)
|
@sebastien-boulle thanks for your pr, I've fixed the issue in this pr: #1267 |
Description
This PR updates the Gemini configuration documentation to provide a out-of-the-box working configuration.
Changes
This also fixed an issue where the previous configuration used "gemini/text-embedding-004" while it was "text-embedding-004" in the pipeline.
References:
Based on Google's Gemini API documentation
Summary by CodeRabbit