This repository has been archived by the owner on Jul 20, 2024. It is now read-only.
v2.1.0 Tarawih 🕌
2.1.0 (2024-03-29)
Features
- add anthropic chat api (66b2785)
- add api registry and modify ToolkitAPI to register (b6e794f)
- add arxiv articles lookup api (f21c64b)
- add auth to write metadata request (1e2cbb5)
- add auth to write metadata request (4373f6e)
- add background_image_removal tool (f73a873)
- add background_image_removal tool (cf65ccb)
- add bing_search api (9607efc)
- add bm25 retrieval tool (bd9b02f)
- add cohere chat api (231c1f6)
- add cohere text embedding (d35e28a)
- Add duckduckgo_search Tool (6bc28e6)
- add duckduckgo_search api (c3f9559)
- add easyocr tool (afd2be0)
- add embeddings_filter tool (262414f)
- add flashrank re ranking tool (e21602f)
- add gemini text embedding (64f1862)
- add gemini text generation api (1da15aa)
- add gif type to video_converter and hykosdk (556d71a)
- add google search api (88cb7a1)
- add groq chat api (e20a85d)
- add huggingface chat api (b797c48)
- add hybrid_search tool (0ca00a6)
- add hyko_toolkit including APIs (30bd596)
- add hyko_toolkit including APIs (16a5a12)
- add image_converter tool (6717b24)
- add image_to_image_with_a_mask api (a0c228a)
- add image_to_image_with_prompt api (425b7b9)
- add image_upscaler stability_ai api (c72d68a)
- add markdown_to_pdf tool (b2f7063)
- add markdown_to_pdf tool (cb75d7f)
- add mask-generation pipeline (1a14530)
- add mask-generation pipeline (6a6bb8c)
- add new video types (MKV, MOV, AVI, MP4, WMV, and WebM) to hykosdk io (73ae807)
- add new youtube_transcript tool (02160d1)
- Add ocr-pdf-loader Tool to close issue #26 (b18aa75)
- add openai embedding model (cbd075f)
- add openrouter chat api (60105f5)
- add pdf_downloader tool (40ebf96)
- Add pdf_loader Tool (b9ff7a3)
- add plotting_functions tool (1c3e19d)
- Add prompt_former Tool (7e64a3d)
- Add recursive_character_text_splitter Tool (880f069)
- Add remove_special_characters tool (7dd253f)
- add remove_stopwords tool (c03f0d5)
- Add semantic_chunker Tool (6ab9d0f)
- Add similarity_search Tool (c793dea)
- add surya_ocr tool (63f0524)
- add tesseract-ocr tool (69f19ad)
- add text_to_image stability_ai api (bc6a259)
- add UltraLytics Image Classification (fc768b3)
- add UltraLytics Image Classification (b6a7564)
- add ultralytics image instance segmentation (dc979ad)
- add ultralytics image instance segmentation (dedb5eb)
- add ultralytics image object detection (d078fc4)
- add ultralytics image object detection (245a50d)
- add UltraLytics Image Oriented Bounding Boxes Object Detection (17a9c4e)
- add UltraLytics Image Oriented Bounding Boxes Object Detection (95f3c76)
- add UltraLytics Image Pose Estimation (5af9e90)
- add UltraLytics Image Pose Estimation (16f6661)
- add ultralytics Video instance segmentation (187ef87)
- add ultralytics Video instance segmentation (ef6d3c3)
- add ultralytics Video object detection (1f3b5bf)
- add ultralytics Video object detection (065af59)
- add UltraLytics Video Oriented Bounding Boxes Object Detection (a06e57b)
- add UltraLytics Video Oriented Bounding Boxes Object Detection (44cf4c9)
- add UltraLytics Video Pose Estimation (bba4a3f)
- add UltraLytics Video Pose Estimation (e47ef54)
- add video_converter tool (d205b75)
- add wikipedia text lookup api (63ee5c4)
- Add wikipedia_search Tool (6a6d23d)
- add yolo small and xlarge to UltraLytics pipes (4ca3e6a)
- add yolo small and xlarge to UltraLytics pipes (a265e5d)
- add youtube_downloader tool (dcce879)
- Add youtube_transcripts Tool (5cee596)
- added --base and --push flags to builder and updated makefile (6182c9f)
- added 'padding' utility function (1c8961f)
- added absolute dockerfile path to deploy function (33806f1)
- added medium and tiny dockerfiles and updated tookit-builder (b54b921)
- added medium and tiny dockerfiles and updated tookit-builder (5686ceb)
- added web scraping-function (585960e)
- added web scraping-function (4f4d425)
- include category dockerfile, remove threading option, use os path utils (1708b66)
- login to docker registry before building (dac0235)
- login to docker registry before building (a658f00)
- migrated image_converter to use new dockerfile (606c003)
- migrated image_converter to use new dockerfile (9b114b3)
- move dockerfiles to deps seperation (aa0072b)
- move huggingface nlp to new dockerfile format (abbd304)
- move validation to seperate io classes, save both preview and original (e63ce33)
- register openai apis (99f5d45)
- Separate document results with URLs (3b1ebd5)
- seperate deployement of each Hyko toolkit defintion (d51245d)
- seperate deployement of each Hyko toolkit defintion (7d2ae3a)
- update to new makefile (3364a1a)
- update to new makefile (8d4490a)
- updated all toolkit to use new base images names (9ce1bca)
- updated converters/utils dockerfiles to use new ones (4ce8e16)
- updated converters/utils dockerfiles to use new ones (f7368b8)
- updated sdk branch and remove login to registry (8b3d85c)
- use deploy function in toolkit-builder (b0982a2)
- use deploy function in toolkit-builder (4a789a4)
- use generic types for metadata (6fe0eca)
- use pil instead of cv2 (0df4b2d)
- use pil instead of cv2 (86c79fc)
- use pil instead of cv2 (437121d)
- use pil instead of cv2 (7f42fb1)
- use pil instead of cv2 (31f81ab)
- use pil instead of cv2 (a8087a9)
- write metadata to db, rename script to toolkit-builder.py (07e0be7)
- write metadata to db, rename script to toolkit-builder.py (2f560ee)
Bug Fixes
- add admin prefix to env vars to avoid conflict with os username (8c28a54)
- add admin prefix to env vars to avoid conflict with os username (5a8c78e)
- add cache to pip install (b1954fa)
- add missing dep to T5 model (e00b9d7)
- add missing dep to T5 model (453d65f)
- add missing dep to text-to-image (f7b4390)
- add missing dep to text-to-image (36cf512)
- add more enum values and add model to api handler (e458ecd)
- add space X) (8a2726c)
- added .hykoignore to image apis (14ec0fc)
- added min_length param to summarization task (db135c8)
- allow to re register api (e92e5a8)
- copy venv command (a3d83fb)
- docker file path should be outside the for loop (e949c03)
- docker file path should be outside the for loop (9b9452a)
- dockerfile still copying hyko_sdk and toolkit-builder -> toolkit_builder (473615f)
- easier names for float string and int (5147310)
- fix duckduckgo_search uvloop.Loop (0591aaf)
- fix image converter (add encoder 'imencode') (f74dcbb)
- fix validation errors for cohere response (39e8b94)
- fixed enum issue for frontend compatibility (0640a2a)
- fixed enum issue in cohere text_embedding (7e6931a)
- fixed natural_language_processing, regression and similarity_search (007c54f)
- fixed prefix_items and any_of in sdk builder script (42ddc6e)
- fixed Supported models enum names (27aef64)
- go back to func name as its needed for sdk builder (93531c4)
- hot: removed duplicate params in fill-mask and question asnwering (06584ef)
- hot: removed duplicate params in fill-mask and question asnwering (328bbf8)
- hyko_base_type to have string json schema and remove allOf and filename from model schema (622409c)
- image and negative prompt should be optional (3407ace)
- import from models not metadata (aac7d3d)
- import from models not metadata (51e33f7)
- include torch dep in huggingface (067f41c)
- missplaced api metadata (ab7967c)
- move converter to converters category (e77348e)
- move converter to converters category (65a88fa)
- move converters to new definition (b6e765e)
- move converters to new definition (10551bc)
- move image utils to new definition (8689781)
- move image utils to new definition (a904480)
- move nlp model to new definition (1bcb3ad)
- move nlp model to new definition (e083a75)
- move text utils to new function definition (f0f0a2d)
- move text utils to new function definition (a7651d1)
- move to computer_vision (ff0357a)
- move to computer_vision (573e044)
- move to new definition (6ecf7f5)
- move to new definition (a7c4971)
- move to new definition (0c8e25c)
- move to new definition (2cbcc26)
- move to new definition (c10f48f)
- move to new definition (6b44d15)
- move to new definition (86d0a19)
- move to new definition (8637a84)
- move to new definition, fix import (0d199ad)
- move to new definition, fix import (dd47fd9)
- move to new model definition (5b2f5c6)
- move to new model definition (32fda2f)
- move utils to new function definition (36c557e)
- move utils to new function definition (4b476b1)
- multi line apt install command in dockerfile (4554782)
- order and typo in docker files (a92f30a)
- output description (6c322de)
- output list of strings instead of str (935e067)
- pdf downloader, move it to downloaders and change base image (a7313c7)
- remove metadata and use alias for json schema (bc186e2)
- remove default value for api key (3cc828a)
- remove empty lines (f1c2f99)
- remove empty startup params (4c0a0f9)
- remove fill mask extra params (e5a1855)
- remove fill mask extra params (e0d56db)
- remove lstrip from path when saving to db (2b2498e)
- remove print and use relative import inside subprocess call (8aec796)
- remove print and use relative import inside subprocess call (0ba65f9)
- remove space (0af3b87)
- remove space (ad63820)
- remove space (0dbef6d)
- remove spaces (58a2d14)
- rename mediawiki -> wikimedia (f369a74)
- rename mime type to ext (9d70f45)
- rename models from - -> _ (a52bfff)
- rename models from - -> _ (83f8fbe)
- require only metadata on build (a685adb)
- return a list of str, update description (200f714)
- simple pdf to string no multithreading (9b73490)
- spaces in CMD (c78b16d)
- spaces in CMD (9d4fc4d)
- speech recognition model, drop audio channels (a29b420)
- startup params initialized as none (20c92c4)
- startup params initialized as none (2c7e428)
- storage path, sorry (fae4e7f)
- toolkit-builder -> toolkit_builder (ffb53f8)
- toolkit-builder -> toolkit_builder (4a3a3de)
- typo in metadata, each defintion should implement dump_metadata (b8f1c82)
- typo in metadata, each defintion should implement dump_metadata (b4d3cd0)
- udpate dockerfiles not install or copy hyko_sdk (1eecfd1)
- udpate dockerfiles not install or copy hyko_sdk (e663126)
- undo relative imports (2f01bc7)
- undo relative imports (99696a6)
- update description (c8c96e3)
- update metadata descriptions (52c6431)
- update models enum (f1fadd4)
- update storage path to /storage (da923ed)
- use CoreModel for typing (c0208e1)
- video-to-audio no filename passing (b15f498)
- wikimedia import (2d5085f)
Performance Improvements
- mount cache, and use --no-cache with poetry (b3bdcfb)
- mount pip cache for each pip install command (6a5bc04)
- move import httpx inside write function (5bdcfca)
- move import httpx inside write function (16c21b1)
- no need to install toolkit package in sdk image (b3a26a1)
- remove model install from docker file (962c4ba)
- use hyko sdk tiny (bb208b5)
- use hyko-sdk-medium when possible, seperate transformers from torch (f7cbd79)
- use hyko-sdk-medium when possible, seperate transformers from torch (9d3d1b3)
- validate list of returned urls (03123c5)