|
|
213d39f59f
|
doc: update doc after removing op desc
|
2022-11-02 19:34:13 -07:00 |
|
|
|
11063a447b
|
doc: update dw post-op description
|
2022-11-01 23:41:34 -07:00 |
|
|
|
42850074d0
|
doc: add graph build and basics
|
2022-10-29 00:51:30 +08:00 |
|
|
|
c519a58aa1
|
doc: update doc pictures
|
2022-10-28 08:46:54 -07:00 |
|
|
|
643f9c52d2
|
doc: fix various doc issues
|
2022-10-03 19:30:22 -07:00 |
|
|
|
88e4b917e2
|
api: remove append_dw_k3sXp1 APIs
|
2022-10-03 10:57:58 -07:00 |
|
|
|
0ddd8ae794
|
doc: add f16 cpu support documentation
|
2022-08-17 17:53:18 -07:00 |
|
|
|
2b923d7207
|
api: depthwise post-op with any kernel, stride and padding values
|
2022-07-05 19:41:38 -07:00 |
|
|
|
bcba5e1a50
|
doc: adjust wording for f16 accumulation type
|
2022-06-08 11:27:25 -07:00 |
|
|
|
93584a01af
|
api: rename f19 to tf32 to align with HW naming
|
2022-05-19 21:05:47 +02:00 |
|
|
|
e2f4991c54
|
doc: change f16 accumulation data type to f32
|
2022-05-18 16:33:21 -07:00 |
|
|
|
b01f0b5053
|
doc: fix a link
|
2022-05-04 21:02:49 -07:00 |
|
|
|
7f69fa7665
|
doc: fix use of warning tags
|
2022-05-04 21:02:49 -07:00 |
|
|
|
2e4b14a3c0
|
include, src: add f19 fpmath_mode
|
2022-05-04 13:19:26 -07:00 |
|
|
|
6dab33b1d0
|
all: add f64 GPU conv support. (#2625)
|
2022-04-01 17:44:44 -05:00 |
|
|
|
b7275b54b5
|
doc: add a note about dw conv post-op
|
2022-01-21 15:40:26 -08:00 |
|
|
|
06f0392a5a
|
doc: spelling
|
2021-12-29 20:23:54 -08:00 |
|
|
|
249ef81aa8
|
doc: fix a few warnings in doc build
Fix warnings about missing parameters, incorrect parameter names, and
invalid markdown syntax.
|
2021-12-29 20:23:54 -08:00 |
|
|
|
67f73dc870
|
src: add compatibility support for ONEDNN build options and env vars
|
2021-11-02 17:39:28 -07:00 |
|
|
|
6f90d3ade6
|
doc: fix link to prelu subsection (#1926)
|
2021-10-25 10:46:23 -07:00 |
|
|
|
54e7a0673d
|
doc: updated the list of supported hardware
|
2021-09-21 15:08:59 -07:00 |
|
|
|
c6d05294c1
|
cpu: aarch64: add fastmath support from Compute Library
|
2021-09-20 10:47:27 +02:00 |
|
|
|
f21b81bf3e
|
doc, src: fpmath_mode: align env variable name
|
2021-09-10 14:26:05 -07:00 |
|
|
|
987dd76244
|
doc: updated name for Xe Architecture graphics
|
2021-08-26 07:45:37 -07:00 |
|
|
|
729e848b2f
|
api: add prelu post operation
|
2021-08-22 18:02:34 -07:00 |
|
|
|
e556b06c2a
|
doc: clarify the lib numerical behavior, and fpmath_mode
|
2021-08-18 21:59:56 -07:00 |
|
|
|
9a753a2b62
|
cpu: matmul: support sum with different data type
|
2021-08-10 15:23:35 +02:00 |
|
|
|
cd4fb52bcf
|
doc: remove ambiguities from binary post-op
|
2021-07-12 09:36:27 +02:00 |
|
|
|
d60a1bc793
|
api: add binary per_tensor post-op support with tag::any
|
2021-06-09 15:58:22 -07:00 |
|
|
|
1b1e048148
|
api: support sum with zero point
|
2021-06-09 13:24:47 +02:00 |
|
|
|
95b170f50f
|
doc: fix style issues in sum postop description
|
2021-05-20 10:20:18 -07:00 |
|
|
|
d008807bd1
|
doc: explain API for float->int8 conv
|
2021-02-19 14:44:20 -08:00 |
|
|
|
bc1931bb07
|
doc: remove mention of functors in basic concepts
|
2021-01-26 10:49:05 -08:00 |
|
|
|
1236e1d16a
|
doc: replaced the old drupal versions of idz urls with the aem versions
|
2020-08-26 15:27:42 -07:00 |
|
|
|
db23af2edd
|
api: add binary post operation
|
2020-08-13 17:01:46 -07:00 |
|
|
|
3bce1bc265
|
examples: gpu: enable bf16 cnn training example
|
2020-06-17 15:05:47 +00:00 |
|
|
|
bc4404427c
|
doc: doesn't -> does not
|
2020-06-16 16:38:09 -07:00 |
|
|
|
b44224db5c
|
src: gpu: conv: add support for different sum and dst data types
|
2020-06-13 01:09:38 +02:00 |
|
|
|
4385841723
|
doc: various updates based on spec work
|
2020-06-02 15:39:49 -07:00 |
|
|
|
86b9a0534e
|
doc: add more details about Xe architecture support
|
2020-05-29 13:33:16 -07:00 |
|
|
|
c5ab293681
|
doc: added Xe architecture support to documentation
|
2020-05-20 18:23:24 -07:00 |
|
|
|
c1728236ca
|
doc: add DG1 and Tiger Lake support
|
2020-05-20 16:26:49 -07:00 |
|
|
|
a0d55c740f
|
doc: add performance settings article
Co-authored-by: Roman Dubtsov <roman.s.dubtsov@intel.com>
|
2020-04-19 20:09:42 -07:00 |
|
|
|
17685bd047
|
doc: cleaned up trademarks use
|
2020-04-08 09:44:04 -07:00 |
|
|
|
00ab48838e
|
Remove "MKL" from images in the documentation
|
2020-04-07 08:42:49 -07:00 |
|
|
|
a30dbbdd14
|
doc: updated name to oneDNN
|
2020-04-03 12:49:00 -07:00 |
|
|
|
818e0315d5
|
doc: cleaned up use of Intel MKL-DNN
|
2020-04-03 12:49:00 -07:00 |
|
|
|
816cb33bfb
|
doc: dnnl programming model diagram and introduction
|
2020-03-13 10:23:27 -07:00 |
|
|
|
87f98e00bc
|
doc: Depthwise Post-op
|
2020-03-04 22:59:55 -08:00 |
|
|
|
6934076ca6
|
doc: reformat tensor names in math mode
|
2020-02-14 16:00:35 -08:00 |
|
