Stefano Fiorucci
f43bc562d3
refactor: replace torch.no_grad with torch.inference_mode (where possible) ( #3601 )
...
* try to replace torch.no_grad
* revert erroneous change
* revert other module breaking
* revert training/base
2022-11-23 09:26:11 +01:00
Stefano Fiorucci
3040e59c63
feat: add support for BM25Retriever in InMemoryDocumentStore ( #3561 )
...
* very first draft
* implement query and query_batch
* add more bm25 parameters
* add rank_bm25 dependency
* fix mypy
* remove tokenizer callable parameter
* remove unused import
* only json serializable attributes
* try to fix: pylint too-many-public-methods / R0904
* bm25 attribute always present
* convert errors into warnings to make the tutorial 1 work
* add docstrings; tests
* try to make tests run
* better docstrings; revert not running tests
* some suggestions from review
* rename elasticsearch retriever as bm25 in tests; try to test memory_bm25
* exclude tests with filters
* change elasticsearch to bm25 retriever in test_summarizer
* add tests
* try to improve tests
* better type hint
* adapt test_table_text_retriever_embedding
* handle non-textual docs
* query only textual documents
2022-11-22 09:24:52 +01:00
tstadel
0d45cbce56
convert eval metrics to python float ( #3612 )
2022-11-22 09:05:10 +01:00
Espoir Murhabazi
d114a994f1
refactor: update Squad data ( #3513 )
...
* refractor the to_squad data class
* fix the validation label
* refractor the to_squad data class
* fix the validation label
* add the test for the to_label object function
* fix the tests for to_label_objects
* move all the test related to squad data to one file
* remove unused imports
* revert tiny_augmented.json
Co-authored-by: ZanSara <sarazanzo94@gmail.com>
2022-11-21 11:06:14 +01:00
Massimiliano Pippi
ea75e2aab5
feat: store metadata using JSON in SQLDocumentStore ( #3547 )
...
* add warnings
* make the field cachable
* review comment
2022-11-18 08:26:19 +01:00
Massimiliano Pippi
1399681c81
move milvus tests to their own module ( #3596 )
2022-11-17 16:22:02 +01:00
Massimiliano Pippi
6cd0e337d0
refactor: Generate JSON schema when missing ( #3533 )
...
* removed unused script
* print info logs when generating openapi schema
* create json schema only when needed
* fix tests
* Remove leftover
Co-authored-by: ZanSara <sarazanzo94@gmail.com>
2022-11-17 11:09:27 +01:00
Julian Risch
8052632b64
test: add test to check id_hash_keys is not ignored ( #3577 )
2022-11-17 09:25:02 +01:00
Stefano Fiorucci
dc26e6d43e
fix: Flatten DocumentClassifier output in SQLDocumentStore; remove _sql_session_rollback hack in tests ( #3273 )
...
* first draft
* fix
* fix
* move test to test_sql
2022-11-16 12:20:57 +01:00
Massimiliano Pippi
ba75d39029
fix: discard metadata fields if not set in Weaviate ( #3578 )
...
* fix weaviate bug in returning embeddings and setting empty meta fields
* review comment
2022-11-15 22:02:53 +01:00
tstadel
6ce2d296f4
fix: Elasticsearch / OpenSearch brownfield function does not incorporate meta ( #3572 )
...
* fix meta bug
* adjust brownfield test
2022-11-15 12:13:21 +01:00
Stefano Fiorucci
9de56b0283
fix: write metadata to SQL Document Store when duplicate_documents!="overwrite" ( #3548 )
...
* add_all fixes the bug
* improved test
2022-11-15 10:04:04 +01:00
Massimiliano Pippi
6a48ace9b9
BREAKING CHANGE: remove Milvus1DocumentStore along with support for Milvus < 2.x ( #3552 )
...
* remove milvus1
* leftover
* revert deprecation process
2022-11-15 09:54:55 +01:00
Massimiliano Pippi
057a8c0b4f
refactor: Pinecone tests ( #3555 )
...
* add pytest option to unmock pinecone
* first try
* handle missing answer
* fix labels metadata
* more tests
* adapt workflow
* typo
* address review comments
2022-11-14 15:19:15 +01:00
Massimiliano Pippi
4dfddf0d10
refactor: Refactor Weaviate tests ( #3541 )
...
* refactor tests
* fix job
* revert
* revert
* revert
* use latest weaviate
* fix abstract methods signatures
* pass class_name to all the CRUD methods
* finish moving all the tests
* bump weaviate version
* raise, don't pass
2022-11-14 09:57:30 +01:00
Massimiliano Pippi
3319ef6d1c
refactor: refactor FAISS tests ( #3537 )
...
* fix write docs behaviour
* refactor FAISS tests
* do not remove the sqlite db
* try
* remove extra slash
* Apply suggestions from code review
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* review comments
* Update test/document_stores/test_faiss.py
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* review comments
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
2022-11-08 16:37:01 +01:00
Sara Zan
43b24fd1a7
fix: strip whitespaces safely from FARMReader's answers ( #3526 )
...
* remove .strip()
* check for right-side offset
* return the whitespace-cleaned answer
* lstrip, not rstrip :D
* remove int
* left_offset
* slightly refactor reader fixture
* extend test_output
2022-11-08 09:26:47 +01:00
Mayank Jobanputra
794fe5ffa4
bug: didn't clean up model files after running pytest for test_table_text_retriever_training ( #3534 )
...
* Added tmp path to avoid clean up of model files later
2022-11-07 15:07:04 +05:30
Massimiliano Pippi
255072d8d5
refactor: move dC tests to their own module and job ( #3529 )
...
* move dC tests to their own module and job
* restore global var
* revert
2022-11-04 17:05:10 +01:00
Massimiliano Pippi
2bb81331b7
feat: add SQLDocumentStore tests ( #3517 )
...
* port SQL tests
* cleanup document_store_tests.py from sql tests
* leftover
* Update .github/workflows/tests.yml
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
* review comments
* Update test/document_stores/test_base.py
Co-authored-by: bogdankostic <bogdankostic@web.de>
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
Co-authored-by: bogdankostic <bogdankostic@web.de>
2022-11-04 09:24:19 +01:00
Stefano Fiorucci
1a60e21137
refactor: simplify Summarizer, add Document Merger ( #3452 )
...
* remove generate_single_summary
* update schemas
* remove unused import
* fix mypy
* fix mypy
* test: summarizer doesnt change content
* other test correction
* move test_summarizer_translation to test_extractor_translation
* fix test
* first try for doc merger
* reintroduce and deprecate generate_single_summary
* progress in document merger
* document merger!
* mypy, pylint fixes
* use generator
* added test that will fail in 1.12
* adapt to review
* extended deprecation docstring
* Update test/nodes/test_extractor_translation.py
* Update test/nodes/test_summarizer.py
* Update test/nodes/test_summarizer.py
* black
* documents fixture
Co-authored-by: Sara Zan <sarazanzo94@gmail.com>
2022-11-03 16:04:53 +01:00
Sara Zan
f0be78c6a6
bug: remove useless import in conftest.py ( #3362 )
...
* Remove useless milvus import in conftest
* schemas
* schemas
2022-11-02 19:22:24 +05:30
Stefano Fiorucci
4b0894f4c2
fix: support long texts for labels in ElasticsearchDocumentStore ( #3346 )
2022-11-02 11:16:36 +01:00
Sara Zan
bb1d9983b0
refactor: remove YAML save/load methods for subclasses of BaseStandardPipeline ( #3443 )
...
* remove methods & update docstring
* remove irrelevant test
2022-11-02 10:14:33 +01:00
bogdankostic
60224412bc
feat: Add headline extraction to ParsrConverter ( #3488 )
...
* Add headline extraction to ParsrConverter
* Add sample PDF file
* Add test
* Use extract_headlines if set in convert method
* Integrate PR feedback
2022-10-31 19:00:02 +01:00
Massimiliano Pippi
b694c7b5cb
Document Store test refactoring ( #3449 )
...
* add new marker
* start using test hierarchies
* move ES tests into their own class
* refactor test workflow
* job steps
* add more tests
* move more tests
* more tests
* test labels
* add more tests
* Update tests.yml
* Update tests.yml
* fix
* typo
* fix es image tag
* map es ports
* try
* fix
* default port
* remove opensearch from the markers sorcery
* revert
* skip new tests in old jobs
* skip opensearch_faiss
2022-10-31 15:30:14 +01:00
Sebastian
384663981d
Fixed bug in onnx converter for XLMRoberta architecture ( #3470 )
2022-10-28 15:35:53 +02:00
Sebastian
8db7dfb884
refactor: TableReader ( #3456 )
...
* Refactoring table reader
2022-10-26 20:57:28 +02:00
Sebastian
59857cb492
feat: Speed up reader tests ( #3476 )
...
* Use a smaller reader where possible
* Change scope to module of reader to get faster load times
2022-10-26 19:04:18 +02:00
Sara Zan
05c68b6624
feat: add document_store to all BaseRetriever.retrieve() and BaseRetriever.retrieve_batch() implementations ( #3379 )
...
* add document_store to retrieve()]
* mypy & pylint
* pass docstore to embedding encoders
* schemas
* mypy and pylint
* fix tfidfretriever
* pylint
* mypy
* pylint
* fix tfidf
* mypy
* pylint
* schemas
* another fix for tfidf
* fix question generation tests
* remove docstore from embedding encoder signature
* pylint
* revert accidental test changes
* Apply suggestions from code review
* check for docstore similarity function only if the docstore is present
* check for docstore similarity function only if the docstore is present
2022-10-26 15:47:06 +02:00
Julian Risch
d0691a4bd5
bug: replace decorator with counter attribute for pipeline event ( #3462 )
2022-10-26 12:09:04 +02:00
bogdankostic
4fbe80c098
feat: Extraction of headlines in markdown files ( #3445 )
...
* Extract headings from markdown files + adapt PreProcessor
* Add tests
* Fix mypy
* Generate JSON schema
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Update haystack/nodes/file_converter/markdown.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply black
* Add PR feedback
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
2022-10-26 11:57:55 +02:00
Vladimir Blagojevic
5ca96357ff
feat: Add CohereEmbeddingEncoder to EmbeddingRetriever ( #3453 )
2022-10-25 17:52:29 +02:00
Stefano Fiorucci
54ec13eaf7
refactor: Change no_answer attribute ( #3411 )
...
* always run validation
* update schemas
* no_answer as a property. break things!
* forgotten schema
* fix
* update openapi
* removed my unnecessary test
* fix sql document store
Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
2022-10-25 13:07:00 +02:00
Mayank Jobanputra
d48577b4e7
bug: removed duplicated meta "name" field addition to content before embedding in update_embeddings workflow ( #3368 )
...
* Removed explicit passage formatting by name field
* passing correct input type for embedding the docs
* Updated test, updated similarity scores and added results
* changed expected input to embed method
2022-10-25 14:52:05 +05:30
Sara Zan
cbf44413d8
feat: add __cointains__ to Span ( #3446 )
...
* add __contains__
* add tests
2022-10-21 13:58:17 +02:00
Vladimir Blagojevic
8f31228211
feat: Add exponential backoff decorator; apply it to OpenAI requests ( #3398 )
2022-10-19 17:47:38 +02:00
Ursin Brunner
5fedfb03b0
fix: Fix the error of wrong page numbers when documents contain empty pages. ( #3330 )
...
* Fix the error of wrong page numbers when documents contain empty pages.
* Reformat using git hooks.
* Use a more descriptive placeholder
2022-10-18 17:51:02 +02:00
Sebastian
93817f63b4
feat: Speed up integration tests (nodes) ( #3408 )
...
* Changed summarizer model to a smaller one (2GB to 500MB) to save on space and speed up the tests.
* Removed google pegasus from cache
2022-10-18 16:23:57 +02:00
Sebastian
15a59fd040
feat: Updated EntityExtractor to handle long texts and added better postprocessing ( #3154 )
...
* Remove dependence on HuggingFace TokenClassificationPipeline and group all postprocessing functions under one class
* Added copyright notice for HF and deepset to entity file to acknowledge that a lot of the postprocessing parts came from the transformers library.
* Fixed text squishing problem. Added additional unit test for it.
Co-authored-by: ju-gu <julian.gutsch@deepset.ai>
2022-10-17 21:26:44 +02:00
Unai Garay Maestre
3a2c8ae3c5
bug: Adds better way of checking query in BaseRetriever and Pipeline.run() ( #3304 )
...
* changes how query and queries are checked if they have been passed in BaseRetriever
* Fixes checking query properly in Pipeline run
* Fixes checking query properly in Pipeline run
* Adds test for FilterRetriever using run method when query is empty
* Adds mock filter retriever and adapts test
* Removes old test, adds MockRetriever to test file and test uses document_store
* Logs error when query is not of type string with a new test for run batch
* Update test/nodes/test_retriever.py
* schemas
2022-10-17 19:00:13 +02:00
Sara Zan
101d2bc86c
feat: MultiModalRetriever ( #2891 )
...
* Adding Data2VecVision and Data2VecText to the supported models and adapt Tokenizers accordingly
* content_types
* Splitting classes into respective folders
* small changes
* Fix EOF
* eof
* black
* API
* EOF
* whitespace
* api
* improve multimodal similarity processor
* tokenizer -> feature extractor
* Making feature vectors come out of the feature extractor in the similarity head
* embed_queries is now self-sufficient
* couple trivial errors
* Implemented separate language model classes for multimodal inference
* Document embedding seems to work
* removing batch_encode_plus, is deprecated anyway
* Realized the base Data2Vec models are not trained on retrieval tasks
* Issue with the generated embeddings
* Add batching
* Try to fit CLIP in
* Stub of CLIP integration
* Retrieval goes through but returns noise only
* Still working on the scores
* Introduce temporary adapter for CLIP models
* Image retrieval now works with sentence-transformers
* Tidying up the code
* Refactoring is now functional
* Add MPNet to the supported sentence transformers models
* Remove unused classes
* pylint
* docs
* docs
* Remove the method renaming
* mpyp first pass
* docs
* tutorial
* schema
* mypy
* Move devices setup into get_model
* more mypy
* mypy
* pylint
* Move a few params in HaystackModel's init
* make feature extractor work with squadprocessor
* fix feature_extractor_kwargs forwarding
* Forgotten part of the fix
* Revert unrelated ES change
* Revert unrelated memdocstore changes
* comment
* Small corrections
* mypy and pylint
* mypy
* typo
* mypy
* Refactor the call
* mypy
* Do not make FARMReader use the new FeatureExtractor
* mypy
* Detach DPR tests from FeatureExtractor too
* Detach processor tests too
* Add end2end marker
* extract end2end feature extractor tests
* temporary disable feature extraction tests
* Introduce end2end tests for tokenizer tests
* pylint
* Fix model loading from folder in FeatureExtractor
* working o n end2end
* end2end keeps failing
* Restructuring retriever tests
* Restructuring retriever tests
* remove covert_dataset_to_dataloader
* remove comment
* Better check sentence-transformers models
* Use embed_meta_fields properly
* rename passage into document
* Embedding dims can't be found
* Add check for models that support it
* pylint
* Split all retriever tests into suites, running mostly on InMemory only
* fix mypy
* fix tfidf test
* fix weaviate tests
* Parallelize on every docstore
* Fix schema and specify modality in base retriever suite
* tests
* Add first image tests
* remove comment
* Revert to simpler tests
* Update docs/_src/api/api/primitives.md
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Update haystack/modeling/model/multimodal/__init__.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* get_args
* mypy
* Update haystack/modeling/model/multimodal/__init__.py
* Update haystack/modeling/model/multimodal/base.py
* Update haystack/modeling/model/multimodal/base.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Update haystack/modeling/model/multimodal/sentence_transformers.py
* Update haystack/modeling/model/multimodal/sentence_transformers.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Update haystack/modeling/model/multimodal/transformers.py
* Update haystack/modeling/model/multimodal/transformers.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Update haystack/modeling/model/multimodal/transformers.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* Update haystack/nodes/retriever/multimodal/retriever.py
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
* mypy
* mypy
* removing more ContentTypes
* more contentypes
* pylint
* add to __init__
* revert end2end workflow for now
* missing integration markers
* Update haystack/nodes/retriever/multimodal/embedder.py
Co-authored-by: bogdankostic <bogdankostic@web.de>
* review feedback, removing HaystackImageTransformerModel
* review feedback part 2
* mypy & pylint
* mypy
* mypy
* fix multimodal docs also for Pinecone
* add note on internal constants
* Fix pinecone write_documents
* schemas
* keep support for sentence-transformers only
* fix pinecone test
* schemas
* fix pinecone again
* temporarily disable some tests, need to understand if they're still relevant
Co-authored-by: Agnieszka Marzec <97166305+agnieszka-m@users.noreply.github.com>
Co-authored-by: bogdankostic <bogdankostic@web.de>
2022-10-17 18:58:35 +02:00
Vladimir Blagojevic
159cd5a666
feat: Add OpenAIEmbeddingEncoder to EmbeddingRetriever ( #3356 )
2022-10-14 15:01:03 +02:00
Vladimir Blagojevic
5ebe3cb33d
fix: QuestionGenerator generates wrong document questions for non-default num_queries_per_doc parameter ( #3381 )
2022-10-14 12:08:30 +02:00
Stefano Fiorucci
7290196c32
fix: allow same vector_id in different indexes for SQL-based Document stores ( #3383 )
...
* fix_multiple_indexes
* improve test names
2022-10-14 09:55:56 +02:00
Massimiliano Pippi
31fa75e9fd
feat: add support for Elasticsearch 7.16.2 ( #3318 )
...
* bump elastic to 7.16.2+
* decouple Elasticsearch and Opensearch
use method override instead of func variables
fix mypy
default value
fix broken tests
update schema
* relax version pin
* rename the base class
* rename module
* fix import order
* do not run the new tests in the old job
* remove outdated TODO
2022-10-13 11:53:27 +02:00
Sebastian
75641dd024
fix: Added checks for DataParallel and WrappedDataParallel ( #3366 )
...
* Added checks for DataParallel and WrappedDataParallel
* Update isinstance checks according to pylint recommendation
* Using isinstance over types
* Added test for dpr training
2022-10-13 08:05:56 +02:00
Malte Pietsch
fb02b61e90
Update README.md ( #3247 )
2022-10-11 10:43:17 +02:00
tstadel
7fe5003c97
fix: eval() with add_isolated_node_eval=True breaks if no node supports it ( #3347 )
...
* fix isolated eval for pipelines without a node supporting isolated mode
* reformat
* add test
2022-10-10 20:48:13 +02:00
bogdankostic
84aff5e2b3
fix: Allow less restrictive values for parameters in Pipeline configurations ( #3345 )
...
* fix: Allow arbitrary values for parameters in Pipeline configurations
* Add test
* Adapt expected error message in tests
* Fix bug
* Fix bug on checking JSON
* Remove test cases that previously tested if error was thrown
* Change encoding in test
* Restrict possible values
* Re-add tests
* Re-add tests
* Add value flag to list elements
2022-10-10 13:08:45 +02:00
