Sebastian Husch Lee
16fc41cd95
feat: Relax requirement for creating a ToolCallDelta dataclass ( #9582 )
...
* Relax our requirement for ToolCallDelta to better match ChoiceDeltaToolCall and ChoiceDeltaToolCallFunction from OpenAI
* Add reno
* Update tests
2025-07-03 08:50:29 +02:00
Amna Mubashar
9fd552f906
chore: remove deprecated async_executor param from ToolInvoker ( #9571 )
...
* Remove async executor
* Add release notes
* Linting
* update release notes
2025-07-02 14:02:51 +02:00
Amna Mubashar
adb2759d00
chore: remove deprecated State from haystack.dataclasses ( #9578 )
...
* Remove deprecated class
* Remove state from pydocs
2025-07-02 12:19:06 +02:00
Stefano Fiorucci
848115c65e
fix: fix print_streaming_chunk + add tests ( #9579 )
...
* fix: fix print_streaming_chunk + add tests
* rel note
2025-07-01 16:49:18 +02:00
Sebastian Husch Lee
3aaa201ed6
feat: Add tool_invoker_kwargs to Agent ( #9574 )
...
* Add new param to agent to pass any kwargs to tool invoker
* Add reno
2025-07-01 11:09:58 +02:00
Amna Mubashar
f11870b212
fix: adjust the async executor in ToolInvoker ( #9562 )
...
* Fix bug
* Add a new test
* PR comments
* Add another test
* Small fix
* Fix linting
* Update tests
2025-06-30 15:13:01 +02:00
Sebastian Husch Lee
97e72b9693
feat: Add to_dict and from_dict to ByteStream ( #9568 )
...
* Add to_dict and from_dict to ByteStream
* Add reno
* Add unit tests
* Fix and expand tests
* Fix typing
* PR comments
2025-06-30 11:57:22 +00:00
Sebastian Husch Lee
fc64884819
fix: Fix _convert_streaming_chunks_to_chat_message ( #9566 )
...
* Fix conversion
* Add reno
* Add unit test
2025-06-30 11:51:25 +02:00
mathislucka
c54a68ab63
fix: files should not be passed as single string ( #9559 )
...
* fix: files should not be passed as single string
* chore: we want word splitting in this case
* fix: place directive before command
* fix: find correct directive placement
2025-06-27 11:17:42 +02:00
Stefano Fiorucci
c18f81283c
chore: fix deepset_sync.py for pylint + general linting improvements ( #9558 )
...
* chore: fix deepset_sync.py for pylint
* check .github with ruff
* fix
* Update .github/utils/pyproject_to_requirements.py
Co-authored-by: Sebastian Husch Lee <10526848+sjrl@users.noreply.github.com>
---------
Co-authored-by: Sebastian Husch Lee <10526848+sjrl@users.noreply.github.com>
2025-06-27 07:54:22 +00:00
mathislucka
101e9cdc34
docs: sync code to deepset workspace ( #9555 )
...
* docs: sync code to deepset workspace
* fix: naming
* fix: actionlint
2025-06-27 07:51:59 +02:00
Stefano Fiorucci
bcaef53cbc
test: export HF_TOKEN env var in e2e environment ( #9551 )
...
* try to fix e2e tests for private NER models
* explanatory comment
* extend skipif condition
2025-06-25 15:00:28 +02:00
Haystack Bot
85e8493f4f
Update unstable version to 2.16.0-rc0 ( #9554 )
...
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-06-25 14:57:16 +02:00
Amna Mubashar
1cd0a128d0
feat: enable parallel tool execution in ToolInvoker ( #9530 )
...
* Enable parallel tool execution in ToolInvoker
* Update handling of errors
* Small fixes
* Small fixes
* Adapt number of executors
* Add release notes
* Add parallel tool calling to sync run
* Deprecate async_executor
* Deprecate async_executor
* Add thread lock
* extract methods
* Update release notes
* Update release notes
* Updates
* Add new tests
* Add test for async
* PR comments
2025-06-25 13:32:11 +02:00
Vladimir Blagojevic
91094e1038
feat: Add finish_reason field to StreamingChunk ( #9536 )
...
* Initial commit
* Update deprecation version
* Improve comment
* Minor simplification
* Add reno note
* Remove deprecation warning
* Remove fallback in haystack/components/generators/utils.py
* FinishReason alphabetical import
* Add tool_call_results finish reason, adapt codebase
* Define finish_reason to be Optional[FinishReason]
* Add StreamingChunk finish_reason in HF generators
* Update reno note
* Repair merge issue
* Update tests for finish_reason
* Resolve mypy issues
* Lint issue
* Enhance HF finish_reason translation
* Remove irrlevant test
* PR comments
---------
Co-authored-by: Sebastian Husch Lee <sjrl423@gmail.com>
2025-06-25 09:06:01 +00:00
Julian Risch
1d1c13a8bc
chore: add DocusaurusRenderer and use description, title, id ( #9538 )
2025-06-25 09:56:26 +02:00
Stefano Fiorucci
0d0a66b4f5
feat: add LLMMessagesRouter, a component to route Chat Messages using LLMs ( #9540 )
...
* llmmessagesrouter - draft
* serde methods
* refinements, tests and release note
* Apply suggestions from code review
Co-authored-by: Daria Fokina <daria.fokina@deepset.ai>
---------
Co-authored-by: Daria Fokina <daria.fokina@deepset.ai>
2025-06-24 14:54:20 +02:00
Michele Pangrazzi
3207a76d50
chore: Update pydoc-markdown.sh ( #9547 )
...
* Make config path a $1 param ; Add usage in comment ; Add echo log
* Update sync command
2025-06-24 14:01:51 +02:00
Amna Mubashar
9ed0b9b0bc
fix: Update the de/serialization with schema utils ( #9526 )
...
* Update the util methods
* Update tests
* fix tests
* schema fix
* Add json schema for tuples and sets
* Add proper conversion for sets and tuples
* Adjust typing
* PR comments
* Linting
* Optimize deserialization
* remove TODO
* PR comments
* PR comments
* Update tests and deserialization error
* Support legacy deserialization
* Update deprecating warning
* Update test
2025-06-24 13:10:12 +02:00
Stefano Fiorucci
d14f5dca0e
feat: add trust_remote_code parameter to SentenceTransformersSimilarityRanker ( #9546 )
2025-06-24 11:39:59 +02:00
Stefano Fiorucci
556dcc9e46
chore: update transformers test dependency ( #9537 )
2025-06-23 10:26:11 +02:00
Sebastian Husch Lee
ec371387f0
refactor: Update to StreamingChunk, better index setting and change tool_call to tool_calls ( #9525 )
...
* Fixes to setting StreamingChunk.index properly and refactoring tests for conversion
* Make _convert_chat_completion_chunk_to_streaming_chunk a member of OpenAIChatGenerator so we can overwrite it in integrations that inherit from it
* Fixes
* Modify streaming chunk to accept a list of tool call deltas.
* Fix tests
* Fix mypy and update original reno
* Undo change
* Update conversion to return a single streaming chunk
* update to print streaming chunk
* Fix types
* PR comments
2025-06-23 08:14:25 +00:00
Ahmad Zidan
f911459647
feat: add resource name for Haystack Component Datadog spans ( #9337 )
...
* feat: add resource name for Haystack Component Datadog spans
* fest: format resource name
Signed-off-by: Ahmad Zidan <ahmad.zidan@traveloka.com>
* feat: add release notes
Signed-off-by: Ahmad Zidan <ahmad.zidan@traveloka.com>
---------
Signed-off-by: Ahmad Zidan <ahmad.zidan@traveloka.com>
Co-authored-by: Sebastian Husch Lee <10526848+sjrl@users.noreply.github.com>
2025-06-18 09:15:15 +00:00
Sebastian Husch Lee
3784889e5d
fix: Fix Tool and ComponentTool serialization when specifying outputs_to_string ( #9524 )
...
* Fix serialization of outputs_to_string in Tool and ComponentTool
* Add reno
* Fix mypy, simplify logic
* fix pylint
* Fix test
---------
Co-authored-by: David S. Batista <dsbatista@gmail.com>
2025-06-18 11:00:46 +02:00
Stefano Fiorucci
a16ee96003
fix: fix SuperComponent class serialization/deserialization for async Pipelines ( #9527 )
...
* draft
* better test + release note
* improve test
2025-06-18 08:17:52 +00:00
Amna Mubashar
67a8f1249b
chore: update linter configuration for compatibility with latest ruff release ( #9528 )
...
* Fix linting
* Fix linting
* Update error suppression
* Update pre commit
* Update pyproject.toml
2025-06-18 09:53:19 +02:00
Sriniketh J
6198f0cba9
feat: adding support for torch xpu device ( #9470 )
...
* feat: add support for torch xpu device support
* test: xpu based tests ci/cd
* test: add xpu code device support
---------
Co-authored-by: Sebastian Husch Lee <10526848+sjrl@users.noreply.github.com>
Co-authored-by: David S. Batista <dsbatista@gmail.com>
2025-06-17 14:15:19 +02:00
baki gul
7dbac5b3c9
Fixes incorrect ID generation for identical chunks in RecursiveDocumentSplitter ( #9517 )
...
* fix(preprocessor): ensure RecursiveDocumentSplitter generates unique chunk IDs
* fix: update meta handling in RecursiveDocumentSplitter to ensure correct overlap information
---------
Co-authored-by: Michele Pangrazzi <xmikex83@gmail.com>
2025-06-16 21:49:00 +02:00
Stefano Fiorucci
7570f6b769
fix: re-export symbols in __init__.py files ( #9521 )
...
* chore: re-export symbols in __init__.py files
* release note
2025-06-16 16:29:08 +02:00
Sebastian Husch Lee
a1484cb91c
Add unit test ( #9519 )
2025-06-16 13:14:02 +02:00
Sebastian Husch Lee
ba6f5eeb9a
feat: Make PipelineBase().validate_input public ( #9520 )
...
* Make validate_input public
* Add reno
2025-06-16 11:58:28 +02:00
Sebastian Husch Lee
c5027d711c
refactor: Refactor HuggingFaceLocalChatGenerator ( #9455 )
...
* Refactoring to better align run and run_async and reduce duplicate code
* Docstrings and align run and run_async
* More changes
* add missing type
* Refactor async part a bit
* Fix import error
* Fix mypy
2025-06-13 15:38:00 +02:00
Sebastian Husch Lee
379df4ab84
feat: Warn users if Agent is called with only system messages ( #9514 )
...
* Add warning message and raise error in agent run method
* Add tests
* Add reno
* Updates
* Updates
2025-06-13 14:58:50 +02:00
Stefano Fiorucci
580683b79d
chore: improve select_streaming_callback type hints ( #9513 )
2025-06-13 14:24:18 +02:00
Mohammed Abdul Razak Wahab
a28b2851d9
feat: Add async streaming support in HuggingFaceLocalChatGenerator ( #9405 )
...
* feat: Add async streaming support in hugging face generator
* enforce streamingcallback to be async
* refactor
* fix: schedule and await async task in Event Loop
* unenforce typecheck
* add integration test
* After merge fixes:
- fix breaking tests
- added component_info to AsyncHFTokenStreamingHandler
* fix integration test
* refactor: improve async handling in HuggingFaceLocalChatGenerator and update tests
* fix typo
* address review comments
* refactors
* typo
* refactor
2025-06-11 14:50:25 +00:00
Stefano Fiorucci
f8155e1b77
chore: clean up ( #9504 )
2025-06-11 11:05:05 +02:00
Sebastian Husch Lee
54c5057e0b
feat: (and fix) Add enable_streaming_passthrough to ToolInvoker and add missing params to to_dict ( #9498 )
...
* Fixes and tests
* Add reno
* Change variable name
* Add test and fix for passing streaming_callback to a component tool
* Add unit test
* Remove unused import
* Fix reno
2025-06-06 14:16:05 +02:00
Amna Mubashar
1d6a9f652a
fix: serialization of nested ChatMessage in GeneratedAnswerdataclass ( #9497 )
...
* Fix serialization
* small fix
* fix the erros
* Fix tests
* PR comments
2025-06-06 11:46:24 +02:00
Stefano Fiorucci
12665ade14
chore: simplify Haystack Hatch scripts ( #9491 )
...
* try unifying hatch scripts
* formatting
* simplify
* improve contributing guidelines
* fmt-check
2025-06-06 10:43:02 +02:00
Sebastian Husch Lee
b61886b138
feat: Update streaming chunk ( #9424 )
...
* Start expanding StreamingChunk
* First pass at expanding Streaming Chunk
* Working version!
* Some tweaks and also make ToolInvoker stream a chunk with a finish reason
* Properly update test
* Change to tool_name, remove kw_only since its python 3.10 only and update HuggingFaceAPIChatGenerator to start following new StreamingChunk
* Add reno
* Some cleanup
* Fix unit tests
* Fix mypy and integration test
* Fix pylint
* Start refactoring huggingface local api
* Refactor openai generator and chat generator to reuse util methods
* Did some reorg
* Reusue utility method in HuggingFaceAPI
* Get rid of unneeded default values in tests
* Update conversion of streaming chunks to chat message to not rely on openai dataclass anymore
* Fix tests and loosen check in StreamingChunk post_init
* Fixes
* Fix license header
* Add start and index to HFAPIGenerator
* Fix mypy
* Clean up
* Update haystack/components/generators/utils.py
Co-authored-by: Julian Risch <julian.risch@deepset.ai>
* Update haystack/components/generators/utils.py
Co-authored-by: Julian Risch <julian.risch@deepset.ai>
* Change StreamingChunk.start to only a bool
* PR comments
* Fix unit test
* PR comment
* Fix test
---------
Co-authored-by: Julian Risch <julian.risch@deepset.ai>
2025-06-06 08:17:02 +00:00
Stefano Fiorucci
f85ce19a32
test: replace tool calling model in tests with Qwen2.5-72B-Instruct ( #9500 )
2025-06-06 08:42:46 +02:00
Sebastian Husch Lee
8e21c501df
fix: Fix serialization and deserialization of ConditionalRouter with multiple outputs ( #9490 )
...
* Fix sede of ConditionalRouter with multiple outputs
* Add reno
2025-06-05 15:57:24 +02:00
David S. Batista
715a9f9347
chore: fixing release notes ( #9496 )
2025-06-05 12:36:40 +02:00
David S. Batista
9c2bc666f9
fixing UID colllision on release notes files ( #9495 )
2025-06-05 12:25:10 +02:00
David S. Batista
529a7f5b6a
docs: fixing typo docstring ( #9493 )
2025-06-05 11:43:17 +02:00
Vladimir Blagojevic
b69d261280
chore: Make docstring-parser core dep ( #9477 )
...
* Make docstring-parser core dep
* Add reno note
2025-06-05 11:28:18 +02:00
Vladimir Blagojevic
853a32f8da
feat: Improve ChatMessage _deserialize_content ValueError - make it more LLM friendly ( #9484 )
...
* Improve ChatMessage _deserialize_content ValueError - make it more LLM friendly
* Add unit test
* Add reno note
* Add descriptive ValueError for missing role
* Update haystack/dataclasses/chat_message.py
Co-authored-by: Stefano Fiorucci <stefanofiorucci@gmail.com>
* Update releasenotes/notes/improve-chatmessage-error-messages-llm-agents-a1b2c3d4e5f6g7h8.yaml
Co-authored-by: Stefano Fiorucci <stefanofiorucci@gmail.com>
* Add role check in ChatMessage
* fixes + refinements
---------
Co-authored-by: Stefano Fiorucci <stefanofiorucci@gmail.com>
2025-06-04 15:14:05 +00:00
Sebastian Husch Lee
db359cff40
Add state to agent pydocs ( #9486 )
2025-06-04 14:01:58 +02:00
Sebastian Husch Lee
ff56363db1
fix: In set_output_types check that the decorator @component.output_types is not present on the run_async method ( #9485 )
...
* Fix
* Add reno
2025-06-04 12:17:47 +02:00
Stefano Fiorucci
1e2214a1a0
feat: ChatMessage.to_openai_dict_format - add require_tool_call_ids parameter ( #9481 )
2025-06-03 16:55:13 +02:00
