adai/MemoryBear
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
3,001 Commits 37 Branches 16 Tags
1744b045fbef26da07a5d1019399e9172fc730fd
Commit Graph

18 Commits

Author SHA1 Message Date
Timebomb2018
377ddd2b9b fix(llm): unify JSON output handling across providers and fix tool+json_output compatibility 
- Remove redundant `response_format` injection for VOLCANO provider since it's unsupported; rely on system prompt injection instead
- Extend system prompt JSON injection logic to cover VOLCANO and tool-enabled cases universally
- Simplify model parameter construction by removing redundant `params["model_kwargs"] = model_kwargs` assignments
- Refactor `CompatibleChatOpenAI._get_request_payload` to strip `response_format` when tools are present, avoiding strict validation errors in langchain_openai
- Fix timestamp calculation order in `datetime_tool.py` to avoid integer truncation before multiplication
 2026-04-17 14:19:40 +08:00
Timebomb2018
ccdf7ae81d refactor(model): replace VolcanoChatOpenAI with CompatibleChatOpenAI for unified omni model support 2026-04-16 17:40:30 +08:00
Timebomb2018
7ba0726473 refactor(model): remove mutual exclusion logic between json_output and deep_thinking 2026-04-16 16:36:15 +08:00
Timebomb2018
8c6b65db12 feat(llm): add json_output support for structured LLM responses 2026-04-16 16:27:55 +08:00
Timebomb2018
737858731b fix(core): conditionally apply thinking parameters based on model support 2026-04-15 12:24:11 +08:00
Timebomb2018
e5e6699168 feat(workflow): support nested variable access and DashScope rerank provider 2026-04-10 16:21:49 +08:00
Timebomb2018
4eed393db5 fix(app): 
1. List operation node sub-variable comparison；
2. Non-Dashscope Omni model processing；
3.Handling the issue of disappearing iterative nodes
 2026-04-08 11:11:57 +08:00
Timebomb2018
9a931389ea fix(app): 
1. Import issue handling；
2. embedding model checkout;
3. omni model removes thinking
 2026-04-07 17:15:32 +08:00
Timebomb2018
264183cec2 feat(models): support reasoning_content streaming 2026-04-01 15:47:43 +08:00
Timebomb2018
876c39b1b0 fix(app): 
1. Token consumption of the omni model;
2. Token consumption of the cluster includes sub-agents
 2026-03-30 18:37:09 +08:00
Timebomb2018
54cff5861a feat(model): add volcano model 2026-03-25 11:45:49 +08:00
Eternity
b71bc1f875 feat(multimodel): support multimodal memory display and improve code style 2026-03-13 14:47:56 +08:00
Eternity
16c1cbe24f feat(agent): add input variable validation 2026-03-05 11:17:56 +08:00
Timebomb2018
590ec3a446 feat(model and app): 
1. Increase support for visual models and multimodal models;
2. The application and workflow can input various multimodal files such as images, documents, audio, and videos.
 2026-03-05 09:55:54 +08:00
Mark
3f42ea2c61 [add] bedrock claude support 2026-02-03 12:05:39 +08:00
Ke Sun
667c3393bc refactor(core/models): reorganize imports and add Bedrock connection pooling 
- Introduce BEDROCK_MAX_POOL_CONNECTIONS environment variable (default: 50)
- Introduce BEDROCK_MAX_RETRIES environment variable (default: 2)
 2025-12-29 14:00:03 +08:00
Mark
a4e276ab27 [MODIFY] Code optimization 2025-12-15 14:09:43 +08:00
Ke Sun
c1adc62ec6 feat: Add base project structure with API and web components 2025-12-02 20:28:01 +08:00