vrr/qwen-code
2
0
Fork 0
mirror of https://github.com/QwenLM/qwen-code.git synced 2026-05-04 06:30:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 10

Merge pull request #2411 from QwenLM/fix-default-output-limit

Browse source 
Increase DEFAULT_OUTPUT_TOKEN_LIMIT from 8K to 16K
This commit is contained in:
Mingholy 2026-03-16 17:34:22 +08:00 • committed by GitHub
commit b4b0041a34
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
2 changed files with 4 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

4
packages/core/src/core/openaiContentGenerator/provider/dashscope.test.ts
View file

@ -817,12 +817,12 @@ describe('DashScopeOpenAICompatibleProvider', () => {
const request: OpenAI.Chat.ChatCompletionCreateParams = { const request: OpenAI.Chat.ChatCompletionCreateParams = {
model: 'unknown-model', model: 'unknown-model',
messages: [{ role: 'user', content: 'Hello' }], messages: [{ role: 'user', content: 'Hello' }],
max_tokens: 10000, // Exceeds the default limit max_tokens: 20000, // Exceeds the default limit
}; };
const result = provider.buildRequest(request, 'test-prompt-id'); const result = provider.buildRequest(request, 'test-prompt-id');
expect(result.max_tokens).toBe(8192); // Should be limited to default output limit (8K) expect(result.max_tokens).toBe(16384); // Should be limited to default output limit (16K)
}); });
it('should preserve other request parameters when limiting max_tokens', () => { it('should preserve other request parameters when limiting max_tokens', () => {

3
packages/core/src/core/tokenLimits.ts
View file

@ -9,7 +9,7 @@ type TokenCount = number;
export type TokenLimitType = 'input' | 'output'; export type TokenLimitType = 'input' | 'output';
export const DEFAULT_TOKEN_LIMIT: TokenCount = 131_072; // 128K (power-of-two) export const DEFAULT_TOKEN_LIMIT: TokenCount = 131_072; // 128K (power-of-two)
export const DEFAULT_OUTPUT_TOKEN_LIMIT: TokenCount = 8_192; // 8K tokens export const DEFAULT_OUTPUT_TOKEN_LIMIT: TokenCount = 16_384; // 16K tokens
/** /**
* Accurate numeric limits: * Accurate numeric limits:
@ -166,6 +166,7 @@ const OUTPUT_PATTERNS: Array<[RegExp, TokenCount]> = [
[/^qwen3\.5/, LIMITS['64k']], [/^qwen3\.5/, LIMITS['64k']],
[/^coder-model$/, LIMITS['64k']], [/^coder-model$/, LIMITS['64k']],
[/^qwen3-max/, LIMITS['64k']], [/^qwen3-max/, LIMITS['64k']],
[/^qwen/, LIMITS['8k']], // Qwen fallback (VL, turbo, plus, etc.): 8K
// DeepSeek // DeepSeek
[/^deepseek-reasoner/, LIMITS['64k']], [/^deepseek-reasoner/, LIMITS['64k']],