a77df06bdd
[+] update payload service
2025-07-25 13:11:44 +03:00
38c9c838ed
[+] let payloads app reuse summarizer
...
1. add rest_client to transform app;
2. wrap summarizer client inside payloads
with domain and protocol configured via settings;
3. update view to request summary generation per each
input list inside POST:/payloads
2025-07-25 13:07:06 +03:00
17f5d81953
[+] deploy payloads app
...
1. add app.py for payloads
alike the one in summarizer app;
2. check that the app service works;
2025-07-25 12:50:04 +03:00
468eac45a2
[+] fix singleton for summarizer
...
1. use explicit singleton class variable
for Summarizer, instead of relying
on fastapi Depends caching;
2025-07-25 12:40:06 +03:00
c0866ee863
[+] fix summarizer created upon every request
...
1. Reuse single Annotated instance;
2025-07-25 12:17:30 +03:00
5568c458c2
[+] improve rest api for summarizer
...
1. refactor out dependency logic into dependencies;
2. fix input schema for summarizer;
3. fix early app initialization before uvicorn sets logging;
4. add manual basicConfig call;
5. update default host and port for fastapi server;
6. add pyright watch mode;
2025-07-25 12:05:20 +03:00
e82586e08c
[+] update service
...
1. adding initialization logging;
2. try to fix no caching for requirements.txt;
2025-07-25 11:35:12 +03:00
520633f383
[+] rework docker for web
...
1. add separate requirements for heavy
packages with custom index-url;
2. rewrite docker file to make sure caching
is not triggering reinstall of requirements
all the time;
3. make sure summarizer has transformers
and torch modules;
2025-07-25 11:23:36 +03:00
491d272304
[+] add app component for transform
...
1. use FastAPI Depends to initialize
Summarizer once for the whole transform app;
2025-07-25 10:40:17 +03:00
d6e1424d89
[+] remove duplicate views from app.py
2025-07-24 11:14:02 +03:00
9bd3c3e45b
[+] partially deploy summarizer as service
...
1. add views, schema for transform module;
2. partially added injection of singleton
instances, like Summarizer;
2025-07-24 11:13:07 +03:00
7355552366
[+] add .whl
2025-07-24 11:03:45 +03:00
64cfccd353
[+] improve worker.py
2025-07-24 11:00:16 +03:00
97e0270550
[+] improve typing for worker.py
2025-07-24 10:26:47 +03:00
c4eb8b5568
[+] partially add worker based on huggingface
...
1. test huggingface manually from IPython;
1.1. seems to work;
1.2 put some long text from documentation,
it has provided some summary;
1.3. runs in about 40-60 seconds on 3-4 cores CPU;
1.4. almost made reuse local cache,
for some reason huggingface still can
download LFS weights from the public repo;
2. partially wrapped into worker
which is to be run as a separate service in docker-compose;
2025-07-23 11:10:13 +03:00
0b5971c4af
[+] partially add transform mock
...
1. researching hugging face model
https://huggingface.co/sshleifer/distilbart-cnn-12-6?library=transformers
2025-07-22 11:14:18 +03:00
80b5a6a64e
[+] update Makefile
2025-07-21 11:25:06 +03:00
bcf300e0c0
[+] update docs, fix url ending
2025-07-21 11:21:44 +03:00
13f5bbc2dd
[+] add return type for rest api payloads
2025-07-21 11:19:59 +03:00
ce50a74510
[+] partially add rest api views
2025-07-21 11:17:26 +03:00
f77399b1d2
[+] add ruff
...
1. add ruff recipies to Makefile;
2. reformat source code with ruff;
2025-07-18 10:11:28 +03:00
207a8737ba
[+] clean async_api
...
1. comment out not used logic;
2. remove app.py and websocket_api.py;
2025-07-18 10:07:43 +03:00
c81ee3c4ec
[+] fix postgresql environment
2025-07-18 10:04:55 +03:00
d185aaf329
[+] add initial payloads/models.py migration
2025-07-18 09:59:16 +03:00
fe33d5c7f6
[+] rename module folder
...
1. rename python module folder;
2. clean up payloads/logic.py;
3. update partially payloads/alembic/env.py;
2025-07-18 09:55:22 +03:00
0ee9e87b7f
[+] update payloads.models
2025-07-18 09:51:05 +03:00
cfdd6b72f0
[+] add requirements
2025-07-17 21:17:41 +03:00
d0ffbeef0f
[+] add simple fastapi template
...
1. rename basic package name;
2. rewrite requirements into docs/readme.md;
2025-07-17 21:14:07 +03:00
