From c1730f00541d444db004fddc2fad5b9e4f241c88 Mon Sep 17 00:00:00 2001
From: SthPhoenix <veloce-raptor@bk.ru>
Date: Sat, 6 Nov 2021 16:48:45 +0300
Subject: [PATCH] Update version to 0.7.0.0

---
 README.md          | 32 ++++++++++++++++++++++++++++++++
 deploy_trt.sh      |  7 +++++--
 src/api_trt/app.py |  2 +-
 3 files changed, 38 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 1707a88..120e5dc 100755
--- a/README.md
+++ b/README.md
@@ -251,6 +251,38 @@ bounding box, detection probability and detection number.
 
 ## Changelog:
 
+### 2021-11-06 v0.7.0.0
+
+Since a lot of updates happened since last release version is updated straight to v0.7.0.0
+
+Additions:
+- Added experimental support for msgpack serializer: helps reduce network traffic for embeddings for ~2x.
+- Output names no longer required for detection models when building TRT engine - correct output order is now extracted 
+  from onnx models.
+- Detection models now can be exported to TRT engine with batch size > 1 - inference code doesn't support it yet, though
+  now they could be used in Triton Inference Server without issues. 
+
+Model Zoo:
+- Added support for WebFace600k based recognition models from InsightFace repo: `w600k_r50` and `w600k_mbf`
+- Added md5 check for models to allow automatic re-download if models have changed.
+- All `scrfd` based models now supports batch dimension/
+
+Improvements:
+- 1.5x-2x faster SCRFD re-implementation with Numba: 4.5 ms. vs 10 ms. for `lumia.jpg` example with
+  `scrfd_10g_gnkps` and threshold = 0.3 (432 faces detected)).
+- Move image normalization step to GPU with help of CuPy (4x lower data transfer from CPU to GPU, about 6% 
+  inference speedup, and some computations offloaded from CPU).
+- 4.5x Faster `face_align.norm_crop` implementation with help of Numba and removal of unused computations.
+  (Cropping 432 faces from `lumia.jpg` example tooks 45 ms. vs 205 ms.).
+- Face crops are now extracted only when needed - when face data or embeddings are requested, improving 
+  detection only performance.
+- Added Numba njit cache to reduce subsequent starts time.
+- Logging timings rounded to ms for better readability.
+- Minor refactoring 
+
+Fixes:
+- Since gender/age estimation model is currently not supported exclude it from models preparing step.
+
 ### 2021-09-09 v0.6.2.0
 
 REST-API
diff --git a/deploy_trt.sh b/deploy_trt.sh
index ed83848..8b3ad99 100755
--- a/deploy_trt.sh
+++ b/deploy_trt.sh
@@ -1,7 +1,7 @@
 #! /bin/bash
 
 IMAGE='insightface-rest'
-TAG='v0.6.3.0'
+TAG='v0.7.0.0'
 
 # Change InsightFace-REST logging level (DEBUG,INFO,WARNING,ERROR)
 log_level=INFO
@@ -42,12 +42,14 @@ force_fp16=False
 det_model=scrfd_10g_gnkps
 
 # REC MODELS:
-## arcface_r100_v1, glintr100
+## arcface_r100_v1, glintr100, w600k_r50, w600k_mbf
 rec_model=glintr100
 ## Do not load recognition model:
 rec_ignore=False
 ## Maximum batch size for recognition model
 rec_batch_size=1
+## Maximum batch size for detection model, use only for building models for Triton Server
+det_batch_size=1
 
 # GENDER/AGE MODELS:
 ## genderage_v1
@@ -109,6 +111,7 @@ for i in $(seq 0 $(($n_gpu - 1)) ); do
         -e REC_NAME=$rec_model\
         -e REC_IGNORE=$rec_ignore\
         -e REC_BATCH_SIZE=$rec_batch_size\
+        -e DET_BATCH_SIZE=$det_batch_size\
         -e GA_NAME=$ga_model\
         -e GA_IGNORE=$ga_ignore\
         -e TRITON_URI=$triton_uri\
diff --git a/src/api_trt/app.py b/src/api_trt/app.py
index a24e79a..630bdf9 100755
--- a/src/api_trt/app.py
+++ b/src/api_trt/app.py
@@ -20,7 +20,7 @@
 from env_parser import EnvConfigs
 from schemas import BodyDraw, BodyExtract
 
-__version__ = "0.6.3.0"
+__version__ = "0.7.0.0"
 
 dir_path = os.path.dirname(os.path.realpath(__file__))