任务池

共 17 条，当前显示第 1–17 条（第 1/1 页）

• core: incorrect token count (`usage_metadata`) in streaming mode

  langchain-ai/langchain·#30429·Python·★ 146784·98 天未动·9 条评论

  buginvestigatecore+1

  66

  可认领
• Empty `AIMessage` when using `astream()` with Anthropic programmatic tool calling

  langchain-ai/langchain·#34406·Python·★ 146784·61 天未动·3 条评论

  buginvestigateanthropic+1

  65

  可认领
• Setting a custom `http_client` fails with unexpected keyword argument when using `ChatAnthropic`

  langchain-ai/langchain·#30146·Python·★ 146784·98 天未动·13 条评论

  buginvestigateexternal

  64

  可认领
• text-splitters: remove `sentence-transformers` dependency

  langchain-ai/langchain·#32336·Python·★ 146784·98 天未动·1 条评论

  investigatetext-splittersinternal

  62

  可认领
• Chunks after `batch_size` treated as non-existant

  langchain-ai/langchain·#32612·Python·★ 146784·44 天未动·5 条评论

  buginvestigatetext-splitters+1

  61

  可认领
• SummarizationMiddleware breaks Anthropic extended thinking by removing thinking blocks during active assistant turns

  langchain-ai/langchain·#34794·Python·★ 146784·39 天未动·4 条评论

  buginvestigatelangchain+1

  61

  可认领
• core: `ToolMessage.status` is not preserved by `core.messages.utils.convert_to_messages`

  langchain-ai/langchain·#32835·Python·★ 146784·31 天未动·5 条评论

  buginvestigatecore+2

  59

  可认领
• `trim_messages` and `ChatAnthropic` token counter with tools

  langchain-ai/langchain·#29637·Python·★ 146784·78 天未动·14 条评论

  buginvestigateintegration+2

  59

  可认领
• core: `_configure_hooks` variable accumulates with each call to `get_usage_metadata_callback`

  langchain-ai/langchain·#32300·Python·★ 146784·30 天未动·5 条评论

  buginvestigatecore+1

  58

  可认领
• String Evaluation with labelled criteria Scoring is wrong

  langchain-ai/langchain·#31870·Python·★ 146784·94 天未动·9 条评论

  help wantedinvestigatelangchain-classic+1

  51

  可认领
• core: leak when using object member in a `RunnableSequence`

  langchain-ai/langchain·#30667·Python·★ 146784·11 天未动·19 条评论

  help wantedbuginvestigate+2

  51

  可认领
• Agent with structured output fails when OpenAI response has phased response with 'commentary' and 'final_answer'

  langchain-ai/langchain·#36290·Python·★ 146784·4 天未动·24 条评论

  buginvestigatelangchain+1

  44

  可认领
• How to apply resources when using create_react_agent?

  langchain-ai/langchain·#40489·Python·★ 146784·5 天未动·16 条评论

  documentationinvestigateexternal+1

  43

  可认领
• Support structured output with Pydantic models in `langchain-huggingface`

  langchain-ai/langchain·#32197·Python·★ 146784·3 天未动·30 条评论

  help wantedinvestigateintegration+2

  40

  可认领
• Reasoning model with structured output

  langchain-ai/langchain·#32120·Python·★ 146784·29 天未动·12 条评论

  help wantedbuginvestigate+3

  37

  可认领
• openai: `bind_tools` not callable after `with_structured_output`

  langchain-ai/langchain·#28848·Python·★ 146784·2 天未动·16 条评论

  help wantedinvestigateopenai+1

  26

  可认领
• Add Perplexity cost tracking

  langchain-ai/langchain·#31647·Python·★ 146784·12 天未动·18 条评论

  help wantedinvestigateintegration+2

  25

  可认领