Stan Kirdey
6038d40a53
Add InMemoryDocumentStore ( #76 )
2020-04-27 21:54:12 +02:00
Tanay Soni
a78659f234
Fix ElasticsearchDocumentStore. get_all_documents() ( #77 )
2020-04-27 16:25:47 +02:00
Tanay Soni
8e736cefa0
Simplify Retriever query ( #73 )
2020-04-27 12:19:59 +02:00
Tanay Soni
2d509078c4
Split Elasticsearch retriever in separate classes ( #72 )
2020-04-23 16:09:53 +02:00
Tanay Soni
dcb5be45b2
Update Python version in Dockerfile-GPU ( #71 )
2020-04-22 19:41:21 +02:00
Tanay Soni
3fe319ab20
Log results in APM ( #70 )
2020-04-22 11:28:23 +02:00
Malte Pietsch
ce4adf3fd2
Update README.rst
2020-04-22 09:55:51 +02:00
Tanay Soni
51a3851f93
Update Dockerfiles to use Gunicorn for deployment ( #69 )
2020-04-21 16:14:51 +02:00
Malte Pietsch
4b98f60c37
fix name of doc offsets in reader
2020-04-21 14:04:46 +02:00
Tanay Soni
f8da804a91
Update num_processes config for FARMReader
2020-04-20 16:09:23 +02:00
Tanay Soni
07df974880
Update FARM version
2020-04-20 15:28:10 +02:00
Malte Pietsch
aafeb2072a
Return more meta data & restructure API reponse format ( #66 )
...
* return more meta data and offsets. move doc name also into meta. change name for feedbackfield from answer_start to answer_doc_start.
* rename offset fields
2020-04-17 19:01:39 +02:00
Malte Pietsch
0e773dfb6d
Update FARM version
2020-04-17 15:56:41 +02:00
Malte Pietsch
c37a685e7f
Update FARM version
2020-04-17 15:55:26 +02:00
Malte Pietsch
21d5a42f7e
Update FARM version
2020-04-17 15:36:24 +02:00
Tanay Soni
7a2e4e7fb7
Udpate FARM version
2020-04-17 15:30:51 +02:00
Tanay Soni
5a93631a3e
Make APM optional ( #65 )
2020-04-17 15:29:19 +02:00
Tanay Soni
8514612153
Update FARMReader config
2020-04-17 15:25:30 +02:00
Malte Pietsch
a4dcfb1fb6
update inferencer args (num_processes, chunksize) to latest FARM version ( #54 )
2020-04-17 15:23:24 +02:00
Tanay Soni
b474d21ced
Add limit on concurrent requests for doc-qa ( #64 )
2020-04-17 15:15:53 +02:00
Malte Pietsch
aa287931f2
add elastic-apm requirement
2020-04-17 14:29:22 +02:00
Malte Pietsch
da4e10ba3c
Update README.rst
2020-04-17 09:44:21 +02:00
Tanay Soni
b5fa084b98
Standardize Finder, Readers, and Retriever interfaces ( #62 )
2020-04-16 13:18:40 +02:00
Tanay Soni
74d9dca261
Modularize API components ( #55 )
...
This PR makes the REST API module more cohesive and maintainable by splitting into separate modules for controllers, routes, config, and the FastAPI app.
2020-04-15 14:04:30 +02:00
Carmen Heger
0cffc6cb1d
Add option for custom elasticsearch queries and filters ( #52 )
...
* added option for custom elasticsearch queries and filters
* switch back to terms query. update docs
Co-authored-by: Malte Pietsch <malte.pietsch@deepset.ai>
2020-04-09 14:18:40 +02:00
Karthik Ganesan
da3277c643
Add document_id with Transformers Reader ( #60 )
2020-04-08 23:27:39 +02:00
Branden Chan
5932aa01c3
Merge pull request #47 from deepset-ai/train_params
...
FARMReader.train() now takes default values from FARMReader
2020-03-24 18:29:26 +01:00
Branden Chan
420e11695b
Remove use_gpu param
2020-03-24 17:47:00 +01:00
Branden Chan
b5c58b32f0
Fix comment
2020-03-24 17:44:08 +01:00
Branden Chan
d41dcca813
FARMReader.train() uses values from class
2020-03-24 17:42:50 +01:00
Malte Pietsch
9897c40c41
Update README.rst
2020-03-23 09:38:55 +01:00
Malte Pietsch
05aa42c687
fix scaling of pseudo probs for es scores. fix filtering of embedding retrieval( #46 )
2020-03-22 18:28:35 +01:00
Malte Pietsch
909ff5d92b
move recently added params for embedding retriever to constructor
2020-03-22 16:29:37 +01:00
Malte Pietsch
456df59586
add params to create_embeddings in retriever ( #45 )
2020-03-22 15:16:30 +01:00
Malte Pietsch
d767f12f7c
Add FAQ-style QA ( #44 )
2020-03-20 16:21:10 +01:00
Malte Pietsch
76c5c1d6aa
Improve deployment of REST API (Configs, logging, minor bugs) ( #40 )
...
* remove env variables from dockerfiles
* add more config options to rest api. make fields optional. change to elasticsearch as default
* skip reader if retriever doesn't return anything
* add more config params to farm reader. fix top_k_per_sample
* update FARM version
2020-03-18 12:26:13 +01:00
bogdankostic
0048ee9c5c
Added Jupyter notebooks of Tutorials ( #43 )
...
Add Jupyter and Colab notebooks of tutorials
2020-03-17 19:58:53 +01:00
Malte Pietsch
51553ef2a1
switch retriever text query from must to should
2020-03-17 07:58:12 +01:00
Malte Pietsch
2164e8550f
Add gpu dockerfile, improve logging, fix minor bug with filtering ( #36 )
...
* add gpu dockerfile. improve logging. fix minor bug with filtering
* fix path
2020-03-12 18:30:42 +01:00
Malte Pietsch
0ddb7c0df6
Add more ES connection params for https ( #35 )
2020-03-03 10:46:59 +01:00
Malte Pietsch
f1a9dad24f
Fix document_name for TFIDF retriever ( #33 )
2020-03-01 16:08:50 +01:00
Malte Pietsch
a0cfd16b0c
fix naming of top_k_per_candidate
2020-02-28 18:27:36 +01:00
Malte Pietsch
b023047bcf
add more granular control over top_k answers within lower level texts (candidates, samples)
2020-02-28 18:24:08 +01:00
Malte Pietsch
eaf42a8c21
upgrade FARM version
2020-02-28 18:23:15 +01:00
Malte Pietsch
eee2676cb0
update docker for fastAPI
2020-02-28 17:49:08 +01:00
Malte Pietsch
30053bb468
Merge branch 'master' of github.com:deepset-ai/haystack
2020-02-28 17:43:38 +01:00
Malte Pietsch
3825d4d797
add basic logging to ElasticsearchRetriever
2020-02-28 17:43:27 +01:00
Aadil Srivastava
f6a3dee2b2
Remove document_name from FARMReader ( #31 )
2020-02-28 17:35:32 +01:00
Tanay Soni
35a22be49b
Add a document_name field in answers ( #30 )
2020-02-27 16:54:50 +01:00
Malte Pietsch
e3db47ff2f
Make Elasticsearch configuration more flexible ( #29 )
2020-02-27 15:11:55 +01:00
