qwen support #2997
Merged
qwen support #2997
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
…schema format issues with smaller models.
Agent Task Evaluation Results: 2/4 (50%)View detailed results
Check the evaluate-tasks job for detailed task execution logs. |
| api_key = os.getenv('ALIBABA_CLOUD') | ||
| base_url = 'https://dashscope-intl.aliyuncs.com/compatible-mode/v1' | ||
|
|
||
| llm = ChatOpenAI(model='qwen-vl-max', api_key=api_key, base_url=base_url) |
There was a problem hiding this comment.
The Qwen model integration in examples/models/qwen.py directly uses the generic ChatOpenAI class, violating the established architectural pattern of creating a dedicated class for each LLM provider (e.g., ChatDeepSeek, ChatGoogle). This inconsistency makes the system harder to maintain and leaks provider-specific configuration like the base_url into example code instead of encapsulating it within a dedicated ChatQwen class.
Prompt for AI agents
Address the following comment on docs/customize/agent/supported-models.mdx at line 226:
<comment>The Qwen model integration in `examples/models/qwen.py` directly uses the generic `ChatOpenAI` class, violating the established architectural pattern of creating a dedicated class for each LLM provider (e.g., `ChatDeepSeek`, `ChatGoogle`). This inconsistency makes the system harder to maintain and leaks provider-specific configuration like the `base_url` into example code instead of encapsulating it within a dedicated `ChatQwen` class.</comment>
<file context>
@@ -208,7 +208,41 @@ llm = ChatOllama(model="llama3.1:8b")
+api_key = os.getenv('ALIBABA_CLOUD')
+base_url = 'https://dashscope-intl.aliyuncs.com/compatible-mode/v1'
+
+llm = ChatOpenAI(model='qwen-vl-max', api_key=api_key, base_url=base_url)
+
+agent = Agent(
</file context>
| "memory": "1-3 sentences of specific memory of this step and overall progress. You should put here everything that will help you track progress in future steps. Like counting pages visited, items found, etc.", | ||
| "next_goal": "State the next immediate goal and action to achieve it, in one clear sentence." | ||
| "action":[{{"one_action_name": {{// action-specific parameter}}}}, // ... more actions in sequence] | ||
| "action":[{{"go_to_url": {{ "url": "url_value"}}}}, // ... more actions in sequence] |
There was a problem hiding this comment.
Example JSON for action contains extra braces and an inline comment, making it invalid JSON. Replace with a valid object structure without comments.
Prompt for AI agents
Address the following comment on browser_use/agent/system_prompt.md at line 212:
<comment>Example JSON for action contains extra braces and an inline comment, making it invalid JSON. Replace with a valid object structure without comments.</comment>
<file context>
@@ -209,7 +209,7 @@ You must ALWAYS respond with a valid JSON in this exact format:
"memory": "1-3 sentences of specific memory of this step and overall progress. You should put here everything that will help you track progress in future steps. Like counting pages visited, items found, etc.",
"next_goal": "State the next immediate goal and action to achieve it, in one clear sentence."
- "action":[{{"one_action_name": {{// action-specific parameter}}}}, // ... more actions in sequence]
+ "action":[{{"go_to_url": {{ "url": "url_value"}}}}, // ... more actions in sequence]
}}
</file context>
Suggested change
| "action":[{{"go_to_url": {{ "url": "url_value"}}}}, // ... more actions in sequence] | |
| "action":[{"go_to_url": { "url": "url_value" }}] |
| await agent.run() | ||
|
|
||
|
|
||
| asyncio.run(main()) |
There was a problem hiding this comment.
Script executes on import; wrap the entrypoint in a main guard to avoid side effects when imported.
Prompt for AI agents
Address the following comment on examples/models/qwen.py at line 26:
<comment>Script executes on import; wrap the entrypoint in a __main__ guard to avoid side effects when imported.</comment>
<file context>
@@ -0,0 +1,26 @@
+ await agent.run()
+
+
+asyncio.run(main())
</file context>
Suggested change
| asyncio.run(main()) | |
| if __name__ == '__main__': asyncio.run(main()) |
| import asyncio | ||
|
|
||
| # get an api key from https://modelstudio.console.alibabacloud.com/?tab=playground#/api-key | ||
| api_key = os.getenv('ALIBABA_CLOUD') |
There was a problem hiding this comment.
No validation for missing API key; running with api_key=None will fail later without a clear, early error.
Prompt for AI agents
Address the following comment on examples/models/qwen.py at line 11:
<comment>No validation for missing API key; running with api_key=None will fail later without a clear, early error.</comment>
<file context>
@@ -0,0 +1,26 @@
+import asyncio
+
+# get an api key from https://modelstudio.console.alibabacloud.com/?tab=playground#/api-key
+api_key = os.getenv('ALIBABA_CLOUD')
+base_url = 'https://dashscope-intl.aliyuncs.com/compatible-mode/v1'
+
</file context>
| {{ | ||
| "memory": "1-3 sentences of specific memory of this step and overall progress. You should put here everything that will help you track progress in future steps. Like counting pages visited, items found, etc.", | ||
| "action":[{{"one_action_name": {{// action-specific parameter}}}}, // ... more actions in sequence] | ||
| "action":[{{"go_to_url": {{ "url": "url_value"}}}}, // ... more actions in sequence] |
There was a problem hiding this comment.
Example JSON contains an inline comment and doubled curly braces, making it invalid JSON despite the instruction to output valid JSON.
Prompt for AI agents
Address the following comment on browser_use/agent/system_prompt_flash.md at line 180:
<comment>Example JSON contains an inline comment and doubled curly braces, making it invalid JSON despite the instruction to output valid JSON.</comment>
<file context>
@@ -177,7 +177,7 @@ You must ALWAYS respond with a valid JSON in this exact format:
{{
"memory": "1-3 sentences of specific memory of this step and overall progress. You should put here everything that will help you track progress in future steps. Like counting pages visited, items found, etc.",
- "action":[{{"one_action_name": {{// action-specific parameter}}}}, // ... more actions in sequence]
+ "action":[{{"go_to_url": {{ "url": "url_value"}}}}, // ... more actions in sequence]
}}
</file context>
Suggested change
| "action":[{{"go_to_url": {{ "url": "url_value"}}}}, // ... more actions in sequence] | |
| "action":[{"go_to_url": { "url": "url_value"}}] |
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.
Auto-generated PR for: qwen support
Summary by cubic
Adds first-class Qwen support via OpenAI-compatible API, with a working example using qwen-vl-max and setup docs. Also tightens action schema examples in prompts to improve output consistency.