feat: auto-detect image aspect ratio from prompt#125
Open
TriTue2011 wants to merge 228 commits into
Open
Conversation
…er, buffered streaming, developer role, tool result context, /v1/responses tool support, usage field)
…nt into last user message
…event hallucination
…use {}, add explicit anti-guessing rule
… hide optional-only params to prevent hallucination
…p_system_hints, _iter_stream_segments, separate role chunk
…model's formatted responses
…from model output
… direct streaming
…riority When search results are detected in user message, move them to the instructions field which Codex treats as mandatory context. This forces the model to use search data instead of asking clarifying questions. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
…earch for external - Keep original user message (don't replace) so tool calling works - Instructions tell model: tools for home info, search for external questions - HA prompt already has 'quote specific numbers from search results' Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
HA prompt already has: 'When search results with real data are provided, always quote specific numbers and facts from them.' Users can customize prompt in HA Conversation agent settings. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- Fixed volume name: chatgpt2api_data with name: for persistence - Added warnings about docker compose down -v - Added troubleshooting section for lost API keys - Added commands to verify and update without data loss Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bind mount to real host directory — data never lost on rebuild. - ./chatgpt2api-data:/app/data - No docker compose down -v risk - Directory visible and verifiable on host Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
…ger combo fallback
… responses - max_tokens was being stripped from Codex requests - Now converts to max_output_tokens (Responses API format) - Default 4096 if not specified by client Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
…back on error - Custom providers in image combo now try adapter (raise if not found) - Combo loop properly tries all models including custom providers - On adapter error, re-raise to trigger next model in combo Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Custom providers that support image gen (like Gemini API server) now have a generic adapter that sends image generation prompt to the chat completions endpoint. Supports: base64 images, image URLs, data URIs in response. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Cho phép tự động nhận diện tỷ lệ khung hình (ví dụ 16:9, 1:1, 4:3) từ nội dung của prompt để ghi đè lên cấu hình mặc định (khi kích thước bị Home Assistant hoặc các client khác gửi lên mặc định). Điều này cho phép ghi đè tham số \size\ bằng những keyword tìm thấy trong yêu cầu gốc của người dùng.