nihui/ncnn - ncnn - 开源协同云脑生态支撑系统

Commit Graph

Author	SHA1	Message	Date
hugo-syn	f35eb4b3b8	chore: Fix multiple typos (#5301 ) Signed-off-by: hugo-syn <hugo.vincent@synacktiv.com>	2 years ago
nihui	a705a24f32	pnnx convert some cudnn conv2d variants (#5289 )	2 years ago
nihui	7ed252c854	pnnx handle index_put with empty indices and scalar values (#5288 )	2 years ago
nihui	c557fb6704	pnnx model binary over 4g (#5274 )	2 years ago
nihui	76dcaa4947	do not eliminate noop math if shape changes, improve torch-2.1 mha attn_mask detection (#5273 )	2 years ago
nihui	6f84952122	pnnx handle more softmin logsoftmax dtype, fuse static full range slices to tensor_split (#5253 ) * pnnx handle more softmin logsoftmax dtype, fuse static full range slices to tensor_split * fix convert nn.Conv2d with none bias tensor * fix embedding input with batch index zero	2 years ago
nihui	c41aa2fdfd	pnnx export with ptpath (#5239 ) * pnnx export with ptpath * build and test python pnnx	2 years ago
nihui	09f2723699	pnnx nn.Identity test (#5238 )	2 years ago
nihui	606626baf3	python pnnx return optimized model (#5232 )	2 years ago
Zhenyu ZHAO	ef29bbec49	Python pnnx with pnnx binary (#5067 )	2 years ago
nihui	3785921e01	pnnx convert nn.Upsample with recompute_scale_factor=True (#5169 )	2 years ago
nihui	4494aadd74	deconvolution dynamic weight (#5119 )	2 years ago
teng	16524ca937	pnnx fix indent errors (#5113 )	2 years ago
nihui	37a108c4c4	pnnx run test_inference by default (#5110 )	2 years ago
nihuini	c2ccbf851d	pnnx fix filename base	2 years ago
nihui	3116e028d2	pnnx drop sdap scale=None for compatiblity with old torch (#5107 )	2 years ago
nihui	14e14a9ae8	slice with indices (#5103 )	2 years ago
nihui	efcfe95a42	pnnx do not sanitize directory part (#5099 )	2 years ago
nihui	0c2a4a29de	pnnx convert batchnorm function (#5097 )	2 years ago
Yoh	3f437d3f3d	Grid sample op (#4373 ) * pnnx support grid_sample op * complete the permute and gridsample operator fusion * spilt calculation into two stages and support permute fusion	2 years ago
佰阅	bcdc276ffe	add torch.view_as_real and torch.view_as_complex (#5083 )	2 years ago
nihui	bedbe599ff	pnnx support torch-2.1 (#5074 )	2 years ago
Marsyule	69d6051850	add F_pairwise_distance to pnnx and ncnn (#4942 )	2 years ago
nihui	78aca88d67	elu 4d and selu 4d (#5047 )	2 years ago
nihui	b8d5a5dbec	pnnx fuse conv3d-bn and deconv3d-bn (#5045 )	2 years ago
nihui	fdf2c482dc	fuse adaptive pool dynamic output size, implement ncnn adaptive pooling dynamic outsize (#5043 )	2 years ago
Amir Ramezani	7e5fa3ade3	shrink operator (#5022 )	2 years ago
Shu Liqiang	a2103be55d	support ReplicationPad3d to ncnn (#4919 )	2 years ago
nihui	f412b7f5c4	fix issue 5025 improve reflect replicated pad pattern	2 years ago
HuPengsheet	6eb994422a	fix build with new protobuf target for pnnx (#5034 )	2 years ago
Shu Liqiang	83cca26794	pnnx convert torch.mv (#4974 )	2 years ago
Amir Ramezani	0ea587b8c7	celu activation vulkan and onnx conversion (#5018 )	2 years ago
Beq Jal	bcfec1da33	Celu layer and export to ncnn (#5019 )	2 years ago
Beq Jal	c851231832	add diag layer and its converter (#4935 )	2 years ago
Amir Ramezani	695f770eab	erf implementation (#5012 ) * added erf implementation * added testcase for erf * added onnx2ncnn support of erf	2 years ago
Jiang-Weibo	4a2620d1e2	pnnx add torch.cumprod op (#5002 )	2 years ago
nicochen1118	4923fb955b	pnnx add torch.positive (#4999 )	2 years ago
莫思潋	379dfd2e29	add torch.lgamma to pnnx (#4976 )	2 years ago
zyt1024	b3fbbccd8b	pnnx convert torch narrow (#4918 )	2 years ago
nihui	b2e94901d2	fix moduleop attribute order, write weight shape as parameter for ncnn module op (#4970 )	2 years ago
nihui	2c587923c0	pnnx fuse moduleop unpack (#4968 )	2 years ago
nihui	cd1c0d6eab	fix build with new protobuf target (#4955 )	2 years ago
nihui	cb674ac5ed	fix build with toolchain defined _L _U constants (#4957 )	2 years ago
nihui	39721eeb94	require c++17 for building with new protobuf (#4947 )	2 years ago
nihui	93e395dc4b	pnnx convert torch maximum minimum and torch max min as expression (#4944 ) * reset device check dtype kind int * placeholder for ncnn sign * convert torch maximum minimum * torch.max as expression * torch.min as expression	2 years ago
lrw04	fed3b43c73	Add logxxx to log comp xxx rewriter where xxx = sigmoid or softmax (#4925 ) * Add logxxx to log comp xxx rewriter * Use pattern matching for LogSigmoid and LogSoftmax * Add conversion passes for functional counterparts * Update documentation	2 years ago
WXB	070a6d40f2	support torch.t to ncnn (#4940 )	2 years ago
nihui	285d0793d4	pnnx fuse expression for scalar-like attribute and unbind chain (#4928 )	2 years ago
nihui	60fedae38b	fix pnnx ghost reshape shape expression inputs, fix intmax overflow on fuse/eval expression (#4923 )	2 years ago
nihui	e02b6e8521	fix pnnx slice copy shape type, inplace op link output (#4914 )	2 years ago

1 2 3 4 5 ...

711 Commits (ff17c170f37d27405f87164fde2271df0c96c8fa)