Merge branch 'main' into ODSC-63450/oci_odsc_llm

.github/workflows/publish_release.yml

@@ -38,7 +38,7 @@ jobs:
         shell: bash
         run: rm -rf llama-index-core/llama_index/core/_static/nltk_cache/corpora/stopwords.zip llama-index-core/llama_index/core/_static/nltk_cache/tokenizers/punkt.zip
       - name: Build and publish to pypi
-        uses: JRubics/poetry-publish@v2.0
+        uses: JRubics/poetry-publish@v2.1
         with:
           python_version: ${{ env.PYTHON_VERSION }}
           pypi_token: ${{ secrets.LLAMA_INDEX_PYPI_TOKEN }}

CHANGELOG.md

@@ -1,5 +1,122 @@
 # ChangeLog
 
+## [2025-01-25]
+
+### `llama-index-core` [0.12.14]
+
+- Fix agentworkflow handoffs for non-openai llms (#17631)
+- small fixes to the multi-agent workflow demo notebook (#17628)
+
+### `llama-index-embeddings-bedrock` [0.5.0]
+
+- Implement async bedrock embeddings (#17610)
+
+### `llama-index-llms-bedrock-converse` [0.4.4]
+
+- Fix prompt stacking in bedrock converse (#17613)
+
+### `llama-index-llms-deepseek` [0.1.0]
+
+- DeepSeek official API LLM (#17625)
+
+### `llama-index-readers-google` [0.6.0]
+
+- GoogleDriveReader support file extensions (#17620)
+
+## [2025-01-23]
+
+### `llama-index-core` [0.12.13]
+
+- Fixing header_path bug re: markdown level vs. stack depth in MarkdownNodeParser (#17602)
+- Advanced text to sql sample rows, adding row retrieval for few-shot prompts (#17479)
+- Made the message role of ReAct observation configurable (#17521)
+- fix reconstructing a tool in AgentWorkflow (#17596)
+- support content blocks in chat templates (#17603)
+- Add contextual retrieval support with a new `DocumentContextExtractor` (#17367)
+
+### `llama-index-graph-stores-memgraph` [0.2.1]
+
+- Vector index support for Memgraph's integration (#17570)
+
+### `llama-index-graph-stores-neo4j` [0.4.6]
+
+- Improves connections for neo4j objects and adds some tests (#17562)
+
+### `llama-index-indices-managed-llama-cloud` [0.6.4]
+
+- Add framework integration for composite retrieval (#17536)
+
+### `llama-index-llms-langchain` [0.5.1]
+
+- get valid string when streaming (#17566)
+
+### `llama-index-llms-mistralai` [0.3.2]
+
+- update function calling models in mistral (#17604)
+
+### `llama-index-llms-openai` [0.3.14]
+
+- fix openai.BadRequestError: Invalid value for 'content': expected a string, got null for tool calls (#17556)
+
+### `llama-index-readers-file` [0.4.3]
+
+- Refactor markdown_to_tups method to better handle multi-level headers (#17508)
+
+### `llama-index-readers-web` [0.3.5]
+
+- feat: Agentql Web Loader (#17575)
+
+### `llama-index-tools-linkup-research` [0.3.0]
+
+- add linkup tool (#17541)
+
+### `llama-index-tools-notion` [0.3.1]
+
+- fix: correct the input params of "load_data" in NotionPageReader (#17529)
+
+### `llama-index-vector-stores-pinecone` [0.4.3]
+
+- build: 🆙 replace pinecone-client with pinecone package (#17587)
+
+### `llama-index-vector-stores-postgres` [0.4.2]
+
+- Add support for halfvec vector type (#17534)
+
+## [2025-01-20]
+
+### `llama-index-core` [0.12.12]
+
+- feat: add AgentWorkflow system to support single and multi-agent workflows (#17237)
+- Fix image-path validation in ImageNode (#17558)
+
+### `llama-index-indices-managed-vectara` [0.4.0]
+
+- (breaking change) API Migration (#17545)
+
+### `llama-index-llms-anthropic` [0.6.4]
+
+- feat: support direct PDF handling for Anthropic (#17506)
+
+### `llama-index-llms-fireworks` [0.3.1]
+
+- Deepseek-v3 is now supported by fireworks (#17518)
+
+### `llama-index-llms-stepfun` [1.0.0]
+
+- feat: add stepfun integrations (#17514)
+
+### `llama-index-multi-modal-llms-gemini` [0.5.0]
+
+- refact: make GeminiMultiModal a thin wrapper around Gemini (#17501)
+
+### `llama-index-postprocessor-longllmlingua` [0.4.0]
+
+- Add longllmlingua2 integration (#17531)
+
+### `llama-index-readers-web` [0.3.4]
+
+- feat: Hyperbrowser Web Reader (#17489)
+
 ## [2025-01-15]
 
 ### `llama-index-core` [0.12.11]

README.md

@@ -37,7 +37,7 @@ LlamaIndex.TS [(Typescript/Javascript)](https://github.com/run-llama/LlamaIndexT
 
 [Documentation](https://docs.llamaindex.ai/en/stable/)
 
-[Twitter](https://twitter.com/llama_index)
+[X (formerly Twitter)](https://x.com/llama_index)
 
 [Discord](https://discord.gg/dGcwcsnxhU)
 

docs/docs/CHANGELOG.md

@@ -1,5 +1,122 @@
 # ChangeLog
 
+## [2025-01-25]
+
+### `llama-index-core` [0.12.14]
+
+- Fix agentworkflow handoffs for non-openai llms (#17631)
+- small fixes to the multi-agent workflow demo notebook (#17628)
+
+### `llama-index-embeddings-bedrock` [0.5.0]
+
+- Implement async bedrock embeddings (#17610)
+
+### `llama-index-llms-bedrock-converse` [0.4.4]
+
+- Fix prompt stacking in bedrock converse (#17613)
+
+### `llama-index-llms-deepseek` [0.1.0]
+
+- DeepSeek official API LLM (#17625)
+
+### `llama-index-readers-google` [0.6.0]
+
+- GoogleDriveReader support file extensions (#17620)
+
+## [2025-01-23]
+
+### `llama-index-core` [0.12.13]
+
+- Fixing header_path bug re: markdown level vs. stack depth in MarkdownNodeParser (#17602)
+- Advanced text to sql sample rows, adding row retrieval for few-shot prompts (#17479)
+- Made the message role of ReAct observation configurable (#17521)
+- fix reconstructing a tool in AgentWorkflow (#17596)
+- support content blocks in chat templates (#17603)
+- Add contextual retrieval support with a new `DocumentContextExtractor` (#17367)
+
+### `llama-index-graph-stores-memgraph` [0.2.1]
+
+- Vector index support for Memgraph's integration (#17570)
+
+### `llama-index-graph-stores-neo4j` [0.4.6]
+
+- Improves connections for neo4j objects and adds some tests (#17562)
+
+### `llama-index-indices-managed-llama-cloud` [0.6.4]
+
+- Add framework integration for composite retrieval (#17536)
+
+### `llama-index-llms-langchain` [0.5.1]
+
+- get valid string when streaming (#17566)
+
+### `llama-index-llms-mistralai` [0.3.2]
+
+- update function calling models in mistral (#17604)
+
+### `llama-index-llms-openai` [0.3.14]
+
+- fix openai.BadRequestError: Invalid value for 'content': expected a string, got null for tool calls (#17556)
+
+### `llama-index-readers-file` [0.4.3]
+
+- Refactor markdown_to_tups method to better handle multi-level headers (#17508)
+
+### `llama-index-readers-web` [0.3.5]
+
+- feat: Agentql Web Loader (#17575)
+
+### `llama-index-tools-linkup-research` [0.3.0]
+
+- add linkup tool (#17541)
+
+### `llama-index-tools-notion` [0.3.1]
+
+- fix: correct the input params of "load_data" in NotionPageReader (#17529)
+
+### `llama-index-vector-stores-pinecone` [0.4.3]
+
+- build: 🆙 replace pinecone-client with pinecone package (#17587)
+
+### `llama-index-vector-stores-postgres` [0.4.2]
+
+- Add support for halfvec vector type (#17534)
+
+## [2025-01-20]
+
+### `llama-index-core` [0.12.12]
+
+- feat: add AgentWorkflow system to support single and multi-agent workflows (#17237)
+- Fix image-path validation in ImageNode (#17558)
+
+### `llama-index-indices-managed-vectara` [0.4.0]
+
+- (breaking change) API Migration (#17545)
+
+### `llama-index-llms-anthropic` [0.6.4]
+
+- feat: support direct PDF handling for Anthropic (#17506)
+
+### `llama-index-llms-fireworks` [0.3.1]
+
+- Deepseek-v3 is now supported by fireworks (#17518)
+
+### `llama-index-llms-stepfun` [1.0.0]
+
+- feat: add stepfun integrations (#17514)
+
+### `llama-index-multi-modal-llms-gemini` [0.5.0]
+
+- refact: make GeminiMultiModal a thin wrapper around Gemini (#17501)
+
+### `llama-index-postprocessor-longllmlingua` [0.4.0]
+
+- Add longllmlingua2 integration (#17531)
+
+### `llama-index-readers-web` [0.3.4]
+
+- feat: Hyperbrowser Web Reader (#17489)
+
 ## [2025-01-15]
 
 ### `llama-index-core` [0.12.11]

docs/docs/api_reference/extractors/documentcontext.md

@@ -0,0 +1,4 @@
+::: llama_index.extractors
+    options:
+      members:
+        - DocumentContextExtractor

docs/docs/api_reference/llms/deepseek.md

@@ -0,0 +1,4 @@
+::: llama_index.llms.deepseek
+    options:
+      members:
+        - DeepSeek

docs/docs/api_reference/tools/linkup_research.md

@@ -0,0 +1,4 @@
+::: llama_index.tools.linkup_research
+    options:
+      members:
+        - LinkupToolSpec

docs/docs/examples/agent/agent_workflow_basic.ipynb

@@ -143,7 +143,7 @@
    "source": [
     "## Maintaining State\n",
     "\n",
-    "By default, the `AgentWorkflow` will maintain statless between runs. This means that the agent will not have any memory of previous runs.\n",
+    "By default, the `AgentWorkflow` will maintain stateless between runs. This means that the agent will not have any memory of previous runs.\n",
     "\n",
     "To maintain state, we need to keep track of the previous state. Since the `AgentWorkflow` is a `Workflow`, the state is stored in the `Context`. This can be passed between runs to maintain state and history."
    ]
@@ -332,7 +332,9 @@
     "\n",
     "\n",
     "async def set_name(ctx: Context, name: str) -> str:\n",
-    "    await ctx.set(\"name\", name)\n",
+    "    state = await ctx.get(\"state\")\n",
+    "    state[\"name\"] = name\n",
+    "    await ctx.set(\"state\", state)\n",
     "    return f\"Name set to {name}\"\n",
     "\n",
     "\n",
@@ -348,8 +350,8 @@
     "response = await workflow.run(user_msg=\"My name is Logan\", ctx=ctx)\n",
     "print(str(response))\n",
     "\n",
-    "name = await ctx.get(\"name\")\n",
-    "print(name)"
+    "state = await ctx.get(\"state\")\n",
+    "print(state[\"name\"])"
    ]
   },
   {