docs: fix multiple documentation warnings - #240

Open

Artlesbol wants to merge 697 commits into

sophgo:masterfrom

Artlesbol:fix-docs-warnings

This pull request is big! We're only showing the most recent 250 commits

Commits on Apr 3, 2025

add bm1684xe

romain-cpu
authored and
PostMalone1998
committed
89ddeed
[vitL] use Lut to accelearte INT8 Softmax

hanqing.zhang
authored and
逸凡陈
committed
a6fa40a
update backend so
dongmingyi
committed
be93e74

Commits on Apr 8, 2025

fix qtable invalid bug

shee-gao
authored and
WhoIsYu
committed
14b4bec
clear mlir_shell code
HarmonyHu
committed
0b5a38e
fix convbwd buffer

liang.wang01
authored and
charlesxzb
committed
1ce49f4
[bugfix] layergroup debug of Float8 case

hanqing.zhang
authored and
charlesxzb
committed
123253b
Revert "add bm1684xe"
PostMalone1998
committed
25eedca

Commits on Apr 9, 2025

refine matmul support_multi_core interface
charlesxzb
committed
072d52d

Commits on Apr 10, 2025

add w8a16 per group quant
romain-cpu
committed
cd72398

Commits on Apr 11, 2025

support to compile llm
HarmonyHu
committed
49e16b3

Commits on Apr 12, 2025

docker info update
charlesxzb
committed
82b0cac
Revise "Optimize the quantization function""
charlesxzb
committed
6ba9bfa

Commits on Apr 13, 2025

fix resnet50 matmul bug
charlesxzb
committed
958b151

Commits on Apr 14, 2025

backend update
charlesxzb
committed
5eba63f
[mars3] fix pass of conv2d

ruiqing.li
authored and
PostMalone1998
committed
c0856c9
[Optimize] correct reshape dtype for PermuteFuseAddSoftmax

seriphdog
authored and
PostMalone1998
committed
91cd8ac
add int4 mix quantization feature

shee-gao
authored and
WhoIsYu
committed
de8e7ae
support to compile llm
HarmonyHu
committed
6edca64
support lmhead and embedding share weights
HarmonyHu
committed
d738a27

Commits on Apr 15, 2025

group_mm perf optimize

xu.yang
authored and
charlesxzb
committed
c83345a
[new] run_calibration support SmoothQuant

wwwwwu64
authored and
WhoIsYu
committed
2fa575f
add auto mixed accuracy quantize for yolo models

wanpeng-zhang
authored and
WhoIsYu
committed
0a0cc94
update backend

romain-cpu
authored and
HarmonyHu
committed
0425c6c
add custom op correlation

xiangzhou.ye
authored and
charlesxzb
committed
0719d96
fix bmodel_dis bug

zwysophon
authored and
charlesxzb
committed
d83e225
[Opt] disable ConvertConv2DToImg2Col for mars3

seriphdog
authored and
PostMalone1998
committed
d7501eb

Commits on Apr 16, 2025

[fix] fix missing scale and zp issue in daily build regressions

wanpeng-zhang
authored and
WhoIsYu
committed
30b343b
support a16matmul

romain-cpu
authored and
HarmonyHu
committed
345ca30
add tpuc-cmd builder
haozhe
committed
fc8d465
fix 1688 w8a16 matmul
HarmonyHu
committed
960e078

Commits on Apr 17, 2025

support bm1684x matmul output int16 meanwhile use perchannel quantization.

wwwwwu64
authored and
WhoIsYu
committed
eb088ce
update docker download url
charlesxzb
committed
ea0622d

Commits on Apr 18, 2025

add coverage build option
haozhe
committed
4fbdf0f
Add int4 and int8 mixed-precision quantization search feature
shee-gao
committed
6abf364
a2 matmul support multi core
HarmonyHu
committed
b653671
[Mars3][FrontEnd] add CastAdd && LNCast && LutMatMul && ActiveCast

seriphdog
authored and
PostMalone1998
committed
e384115
fix matmul efficient
HarmonyHu
committed
edac45a

Commits on Apr 20, 2025

Fixed the issue of models in model-zoo

chang.zhao
authored and
charlesxzb
committed
ed483e1

Commits on Apr 21, 2025

update mix-precsion quantization

shee-gao
authored and
WhoIsYu
committed
9b200f8
[bugfix]correlation calute error in mutil batch
xiangzhou.ye
committed
d00e5bf

Commits on Apr 22, 2025

clear run_qtable related info
charlesxzb
committed
0dbca11
add disable_gdma_check option
charlesxzb
committed
e167beb
[TopOps.td TpuOps.td]

haibiao-yang
authored and
charlesxzb
committed
e02ed7b

Commits on Apr 23, 2025

[tpulang] support qwen2_7b block/block_cache

romain-cpu
authored and
HarmonyHu
committed
bda0c9c
[bugfix][MLIR-484/MLIR-483] SubnetDivide reduce the ops for dynamic
xu.yang
committed
442614e

Commits on Apr 24, 2025

[buffersize] recorrect Conv2d depthwise_waddr for A2 & Mars3 when using 3ic

seriphdog
authored and
PostMalone1998
committed
2363686

Commits on Apr 25, 2025

[LLM] support GPTQ/AWQ quant mode

romain-cpu
authored and
HarmonyHu
committed
cb2b23d
support Qwen2.5VL convert
HarmonyHu
committed
c7f6a58

Commits on Apr 27, 2025

qwen2.5vl support window attention
HarmonyHu
committed
d37da77

Commits on Apr 29, 2025

support qwen3
HarmonyHu
committed
49c5621
[LLM] support sample head convert

sgwzy22
authored and
HarmonyHu
committed
3539903

Commits on Apr 30, 2025

add pattern-match section in quick_start docs

wanpeng-zhang
authored and
WhoIsYu
committed
9d3b6a2
refine readme
HarmonyHu
committed
e621374

Commits on May 6, 2025

[feat] improve profile of 1690
zhou.yang
committed
6e13b4e
feat: LgDebugger & lmem_assign_visualizer
yifan.chen
committed
051d3b5
[Mars3][SGTPUV8][IVA-881] shiyun model build
PostMalone1998
committed
c206d30

Commits on May 7, 2025

[bugfix]calute error for gridsample
xiangzhou.ye
committed
51ffac0

Commits on May 8, 2025

[profile] 1690 add uArch Rate

neocats
authored and
zhou.yang
committed
b7d0655
fix documents
HarmonyHu
committed
6eeb9e0

Commits on May 9, 2025

fix ppl multi-process issue
neocats
committed
0db98bf
[fix] pattern_match YOLO post block ops collection

wanpeng-zhang
authored and
WhoIsYu
committed
057d95e
[tpu] correct active fast_gelu buffer to 3 * tensor_size

seriphdog
authored and
PostMalone1998
committed
aff7a1d

Commits on May 11, 2025

refine README
HarmonyHu
committed
5d8a634
speedup regression for model/script
HarmonyHu
committed
a9ad932

Commits on May 12, 2025

[fix whl bug]

haibiao-yang
authored and
charlesxzb
committed
6a6b0c9
[Bugfix] Fix calculation error in SliceAxis Op

xiangzhou.ye
authored and
逸凡陈
committed
230362a
support instance_norm fp16 compute

xiangzhou.ye
authored and
逸凡陈
committed
2c75b53
fixbug: support output 0231 permute for MatMulHdimBatchPattern
seriphdog
committed
383291c

Commits on May 13, 2025

fix: LG hold_in_lmem bug & update backend
yifan.chen
committed
467410a
modify Appx.02_tpulang_support_op
linhan jia
committed
2d35ccb
[doc] add mini dataset precision test introduction
charlesxzb
committed
4e79c71

Commits on May 14, 2025

1684x and 1688 backend update
xiangzhou.ye
committed
4ce9fbd

Commits on May 15, 2025

speedup regression for op test
HarmonyHu
committed
3588e75
add llm test in regression
HarmonyHu
committed
c9b3627
change description for using tpu-perf
haozhe
committed
3703ea9
refine regression log
HarmonyHu
committed
6c09c61
[feat & fix] support OrOp, fix SliceOp axes

Heng2024
authored and
HarmonyHu
committed
fea1e22

Commits on May 16, 2025

Add a faster version of search_qtable

shee-gao
authored and
WhoIsYu
committed
56be9f2
set max 8 cmodel inference
HarmonyHu
committed
3529110
support automatic generation of openclip qtable during calibration

wwwwwu64
authored and
HarmonyHu
committed
a4c5f51

Commits on May 19, 2025

[Mars3] support board profile
PostMalone1998
committed
5450330
mars3ABCSet-frontend: support softmaxRequant

seriphdog
authored and
PostMalone1998
committed
9e67b35
fix: LG bug of 1684 LutOp weight
yifan.chen
committed
51bc7f1

Commits on May 20, 2025

[fix] negative slice offset

Heng2024
authored and
逸凡陈
committed
5b3a834
refine bmodel regression
HarmonyHu
committed
d668865
[vitl] support lgCostCache
hanqing.zhang
committed
f3337c8

Commits on May 21, 2025

disable invalid local mulshift
charlesxzb
committed
a6b6483

Commits on May 22, 2025

llm_convert support dynamic
HarmonyHu
committed
7c7eb01
update docs for opt_post_processor

seriphdog
authored and
charlesxzb
committed
1123b90
fix clone_bf16 defect
charlesxzb
committed
19cd2d7
style: clang-format -i all files
zhou.yang
committed
ecafc55
[fix] fix the pre-commit
zhou.yang
committed
0c81446
fix A16MatMul dynamic error
HarmonyHu
committed
942faf1
fix the scale calculation error
shee-gao
committed
c46757e

Commits on May 23, 2025

disable concat in 3d group type
charlesxzb
committed
9ed9939
support io_reloc model for 1684x
hanqing.zhang
committed
22ebad6
[fix] mod n == 0 bug

Heng2024
authored and
逸凡陈
committed
98f29a4
add docs for TPU Profile
HarmonyHu
committed
2bfbd1f

Commits on May 26, 2025

llm config file except *.pth *.pt
HarmonyHu
committed
0c3b509

Commits on May 27, 2025

bm168x support yuv preprocess
charlesxzb
committed
5e5f26b
[llm] refine sample head

sgwzy22
authored and
HarmonyHu
committed
4daef75
[mars3] support bf16 for arg

ruiqing.li
authored and
PostMalone1998
committed
0348b85
fix maskrcnn path prob
charlesxzb
committed
5d33b7a
add Code Development in docs
HarmonyHu
committed
33180eb
[doc] v1.19 version info update
charlesxzb
committed
833be2a
[doc] remove tpu_perf install part
charlesxzb
committed
55ee68c

Commits on May 28, 2025

support Gemma3 llm part
HarmonyHu
committed
930d5cf

Commits on May 29, 2025

fix pen mistake
HarmonyHu
committed
989747d
[fix] Slice split inconsistent with merge when dynamic

Heng2024
authored and
逸凡陈
committed
f3404e5

Commits on May 30, 2025

2260 matmul dtype fix
charlesxzb
committed
9c20d4b
[bugfix] Deconv3d

tpu-mlir-vis-tools
authored and
charlesxzb
committed
2f86b38
optimize smoothquant algorithm

wwwwwu64
authored and
WhoIsYu
committed
ae6a366
support automatic generation of bert-l qtable during calibration

wwwwwu64
authored and
WhoIsYu
committed
5533abb

Commits on Jun 4, 2025

[bugfix]Error in conv3d buffer size calculation

xiangzhou.ye
authored and
逸凡陈
committed
251763c
update vit-l pattern in pattern match

wwwwwu64
authored and
WhoIsYu
committed
34e6ceb

Commits on Jun 5, 2025

add pruning pass

hengyang123
authored and
HarmonyHu
committed
631bbd5

Commits on Jun 6, 2025

support smoothing post-norm layernorms in bert

wwwwwu64
authored and
WhoIsYu
committed
69b3bfd
[llm] support InternVl3Converter

sgwzy22
authored and
HarmonyHu
committed
b85b5cf

Commits on Jun 9, 2025

update vit-b and bert patterns pattern match

wwwwwu64
authored and
WhoIsYu
committed
7ccb0a8
[fix] SplitMatMulPattern bug when slice offset < 0

Heng2024
authored and
逸凡陈
committed
df950ea
[fix] add_post_process for yolov11

Heng2024
authored and
逸凡陈
committed
3c27425
[fix] slice backward bug

Heng2024
authored and
逸凡陈
committed
184eef3

Commits on Jun 10, 2025

add shape info for model inference
WhoIsYu
committed
ef7572a
fix qwen2.5vl png in docs
HarmonyHu
committed
522e97a

Commits on Jun 11, 2025

Gemma3 support vision part
HarmonyHu
committed
320f2c6
[Mars3] forbid use dynamic subnet
PostMalone1998
committed
16517bb

Commits on Jun 12, 2025

Add einsum mode

chang.zhao
authored and
逸凡陈
committed
dce3524
[bugfix] Deconv3d int8 & backend update

tpu-mlir-vis-tools
authored and
charlesxzb
committed
724a64e

Commits on Jun 13, 2025

LlmConverter support Qwen2VL
HarmonyHu
committed
93325a2

Commits on Jun 14, 2025

[llm] support MiniCPM4

sgwzy22
authored and
HarmonyHu
committed
b412b2d

Commits on Jun 16, 2025

fix weightop addr assign order bug
charlesxzb
committed
fb30c70
fix TopFuseTile pattern
jianmin.xu
committed
9b33103
[doc] update model-zoo test desc
charlesxzb
committed
9095ac5

Commits on Jun 17, 2025

[Bugfix] fix fattn head for BM1690

zijing.wan
authored and
charlesxzb
committed
2daa316
layergroup: support multi-thread, add new cluster method

hanqing.zhang
authored and
逸凡陈
committed
94e263b
uniform active for llm
HarmonyHu
committed
f56c966
add all ops after the last convs of yolo to qtable

wwwwwu64
authored and
WhoIsYu
committed
b01ad04
yolov5scompile modification
charlesxzb
committed
845f7c2
Revert "support instance_norm fp16 compute"
charlesxzb
committed
c3a905c
format all python code
HarmonyHu
committed
3dd8742

Commits on Jun 18, 2025

set the last convs of yolo to output int16

wwwwwu64
authored and
WhoIsYu
committed
00b7154
[fix] multi inputs when fuse_preprocess
Heng2024
committed
01e6da8

Commits on Jun 23, 2025

[backend update] 2a5e68 -> bcef31
charlesxzb
committed
85d0bbb
TpuLang.py bug fixed:
xu.yang
committed
b41cfd3

Commits on Jun 24, 2025

fix infinite loop in corematch
charlesxzb
committed
f094b51
refine test_fx.py testing case

jianmin.xu
authored and
charlesxzb
committed
7c1dc20
fix: model transform constant fold failed

linhan jia
authored and
charlesxzb
committed
2e34c97
fix: conv bwd 1688 & 1684x

LuTaoChen
authored and
charlesxzb
committed
8bbe5c5
fix pen mistake
HarmonyHu
committed
a70f53d
[sg2260] matmul_fix8b support multi-core

chunlei.zhang
authored and
charlesxzb
committed
d6c7f13

Commits on Jun 25, 2025

add yuv input support

linhan jia
authored and
charlesxzb
committed
4fe8999
[bugfix][MLIR-474] bmodel_combine_coeff failed

xu.yang
authored and
charlesxzb
committed
dc1a029
update release tools
charlesxzb
committed
c096af9

Commits on Jun 26, 2025

add internvl3/qwen2.5vl test
HarmonyHu
committed
4749b41

Commits on Jun 27, 2025

allow perchannel matmul, output int16 matmul and conv in mars3

wwwwwu64
authored and
WhoIsYu
committed
9449696
update 1690 GELU quantize method
charlesxzb
committed
d20c7dc

Commits on Jun 30, 2025

[doc] v1.20 release summary update
charlesxzb
committed
a275c9d
support fp8 for more models
WhoIsYu
committed
7fae13f
llm_convert.py support ChatGLM3
hengyang123
committed
9e6984a

Commits on Jul 1, 2025

backend update
charlesxzb
committed
76c4e96
[fix] WhereBnbwdFusePattern only for bm1690

liang.wang01
authored and
charlesxzb
committed
bba8e91

Commits on Jul 2, 2025

support set max_input for llm
HarmonyHu
committed
5478981

Commits on Jul 3, 2025

[doc] add test notification for QA
charlesxzb
committed
1451ea6
fix layergroup and multi-branch bugs

hanqing.zhang
authored and
逸凡陈
committed
0302fd2
refine debug mode for llm converter
HarmonyHu
committed
965cdf8

Commits on Jul 4, 2025

[Mars3] ArgOp dynamic chosse lowring U16/BF16
PostMalone1998
committed
317af1a
support a special gridsampler and fix nearest mode

xiangzhou.ye
authored and
逸凡陈
committed
7564c1b

Commits on Jul 6, 2025

fix TopReshapeFuse2 loc_name conflict prob
charlesxzb
committed
f151f5c

Commits on Jul 7, 2025

fix matmul gdma_check error
charlesxzb
committed
30c2dca

Commits on Jul 8, 2025

add phi3 converter
hengyang123
committed
3cacfa5

Commits on Jul 11, 2025

add dataset prepare instructions

linhan jia
authored and
charlesxzb
committed
0a778ef

Commits on Jul 14, 2025

[bufix]inplaceOp AddressAssign error

xiangzhou.ye
authored and
逸凡陈
committed
b1a512c

Commits on Jul 15, 2025

fix: local RopeOp broadcast bug
yifan.chen
committed
ac7cb49
[fix] conv slice merge error

Heng2024
authored and
逸凡陈
committed
9c90004
[fix] FusePreprocess bug, when non picture input.shape.size()>=4

Heng2024
authored and
逸凡陈
committed
df3d02c
[feat] disable dynamic layer group when necessary

Heng2024
authored and
逸凡陈
committed
29543b6
[feat] support axis != -1 for TopK

Heng2024
authored and
逸凡陈
committed
4407e14
add time_fixed_subnet pass

xiangzhou.ye
authored and
逸凡陈
committed
22ddd79
support phi-4 converter

hengyang123
authored and
HarmonyHu
committed
25c8cbc

Commits on Jul 16, 2025

[fix] model_runner tpuv7
yiqun.huang
committed
57f07e3

Commits on Jul 17, 2025

[Mars3] update backend
PostMalone1998
committed
6152014
support prefill with kv cache
HarmonyHu
committed
b40b35b

Commits on Jul 18, 2025

fix calibration bug

shee-gao
authored and
WhoIsYu
committed
ca5784b
[llm] add llm_analyse

sgwzy22
authored and
HarmonyHu
committed
2f06262
refine apply_rotary_pos for llm
HarmonyHu
committed
737cdff

Commits on Jul 21, 2025

fix qwen2.5vl dynamic error
HarmonyHu
committed
428b3b2
[mars3] support io_tag

hanqing.zhang
authored and
逸凡陈
committed
ebd1c92
support to set max_prefill_kv_length
HarmonyHu
committed
a0a53d8

Commits on Jul 22, 2025

fix concat addressassign prob
charlesxzb
committed
d73ae07

Commits on Jul 23, 2025

Optimize some Ops can into the layer group
xiangzhou.ye
committed
9d9abec
Add check ref/check diff function in bmodel_checker

jianmin.xu
authored and
charlesxzb
committed
01094fd
disable slice3D group
charlesxzb
committed
65b559b
update inplaceinterface methods
charlesxzb
committed
3da1806

Commits on Jul 24, 2025

feat: add RewriterConfigUtils
yifan.chen
committed
ab81504
[mars3][bringup] workaround for group conv

romain-cpu
authored and
PostMalone1998
committed
abc99eb
[bugfix]: fix coeff location creation
Korbin-chen
committed
48028fb
fix: Deconv Op bias type error when lowering

yifan.chen
authored and
charlesxzb
committed
43ea4ec
set output of matmul to f16 if next op is f16
WhoIsYu
committed
63efb6d

Commits on Jul 25, 2025

[CV184X] mars3 chip rename to cv184x
PostMalone1998
committed
7d65443
extend test_fx function
charlesxzb
committed
35c2dad
feat: refactor bm1684x codgen api

yifan.chen
authored and
charlesxzb
committed
cf39dc7

Commits on Jul 28, 2025

make all unsigned input to add get unsigned result
WhoIsYu
committed
0c7249a
[feat] add op deformable attn

yiqun.huang
authored and
逸凡陈
committed
2f76130

Commits on Jul 29, 2025

fix: bugs of TileOp & ReciprocalOp for dynamic model
yifan.chen
committed
1918f5c
[Bugfix] fix arg coreparell fail
zijing.wan
committed
17babc7

Commits on Jul 31, 2025

add missed modification
WhoIsYu
committed
5ea52cf

Commits on Aug 1, 2025

tpulang add gdma_check control
charlesxzb
committed
d2edd4d
[backend_update] weekly backend update
charlesxzb
committed
7e4ec2a

Commits on Aug 4, 2025

[bugfix] fix convbwd probs
charlesxzb
committed
0037966
[feat] maskrcnn adaptation

Heng2024
authored and
逸凡陈
committed
4c12800
supports dynamic model with ppl
neocats
committed
44c7a19

Commits on Aug 5, 2025

feat: support Qwen2.5-Omni (no talker)
HarmonyHu
committed
2fd4fe7
perf: add 4 tile patterns for BM1688 eva-02 model.
hanqing.zhang
committed
4f9aba6

Commits on Aug 7, 2025

feat: bm1688 support fp16 cast bf16, bf16 cast fp16
zijing.wan
committed
912aa03
fix: fix convbwd type verify bug
charlesxzb
committed
81fb59f
feat: PPL Add local dyn_local sample
neocats
committed
1d82181

Commits on Aug 8, 2025

fix ppl build script
neocats
committed
8e02901
change mem mode to force value mode
WhoIsYu
committed
62a43d4

Commits on Aug 11, 2025

fix: qwen3 multi device parallel error
HarmonyHu
committed
3b74bef

Commits on Aug 12, 2025

perf: layergroup optimize dma w slice
PostMalone1998
committed
015442a
fix: qwen3 multi device parallel error (4/8 dev)
HarmonyHu
committed
6ef2de3

Commits on Aug 13, 2025

chore: update pre-push for commit msg format check
charlesxzb
committed
a50bf8e
fix: fix conv with non-weight bias bug
charlesxzb
committed
338d8db
fix multi-branch group bug for update tensor_infos

hanqing.zhang
authored and
逸凡陈
committed
d92503b
chore: add doc sensitive words checking
charlesxzb
committed
c64a3c8

Commits on Aug 14, 2025

bugfix: fix one multi-thread layergroup bug

hanqing.zhang
authored and
逸凡陈
committed
465db3a
fix: [MLIR-564] compile failed bug fixed
xu.yang
committed
33c3236
refactor: [MLIR-553] tpulang.py get rid of onnxruntime/onnxruntime-extension import request
xu.yang
committed
e50ed1f
fix: [MLIR-513] PermuteReorderPattern bug fix
xu.yang
committed
54f8273
[fix]inplaceOp addr set error

xiangzhou.ye
authored and
逸凡陈
committed
a52ad90

Commits on Aug 15, 2025

fix: dynamic compile bug fix
yifan.chen
committed
a621bb2

Commits on Aug 18, 2025

fix: Fattension dyn HighPrecision
neocats
committed
048c308

Commits on Aug 19, 2025

[fix]: dynamic ConcatOp shape inference

Heng2024
authored and
逸凡陈
committed
2d991bf
chore: rename bmrt_test.py
charlesxzb
committed
5893a70

Commits on Aug 20, 2025

docs: add llm_analyse.py info
HarmonyHu
committed
54431e6
chore: test case adjustment
charlesxzb
committed
b4e0666

Commits on Aug 21, 2025

feat: support SelectiveScanOp

jianmin.xu
authored and
charlesxzb
committed
f23eccf

Commits on Aug 22, 2025

fix: keep i32/8 to si32/8 cast
tpu-mlir-vis-tools
committed
6536123
fix ConvertMatMul2Attention pattern

yubing.shen
authored and
charlesxzb
committed
64f51c6
fix: model_tool show real device mem size

liang.wang01
authored and
charlesxzb
committed
71d8870
feat: add new tpulang ops

liang.wang01
authored and
charlesxzb
committed
6f3b282
feat: support IndexputOp

tpu-mlir-vis-tools
authored and
charlesxzb
committed
0933464
fix: fix bm1688 llama2-7b bmrt_test cmp failed prob
charlesxzb
committed
57ebff3
fix: fix nodechip_interp

chang.zhao
authored and
charlesxzb
committed
f65a88f
chore: tpuc-opt supports set time limit
charlesxzb
committed
d54a962
fix: LayerGroup multi-thread bug

yifan.chen
authored and
charlesxzb
committed
096beef

Commits on Aug 25, 2025

feat: support share prompt
HarmonyHu
committed
ae46d2e
feat: make yolo segment mask to 0-255

jianmin.xu
authored and
charlesxzb
committed
da3c97e
docs: fix multiple documentation warnings
Artlesbol
committed
58d39f1