natas_hw
/
openvino
0
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
openvino/src/plugins
History
Gorokhov Dmitriy b54f753354
[CPU] Fixed BF16 Matmul inference precision (#22995) 
CPU plugin uses EnforceInferencePrecision routine for BF16 precision
mark-up. Its logic assumes only activations precision is changed before
Matmul op, while weights precision keeps w/o any changes. Since
dnnlFCTypeMapping misses BF16 activation, FP32 weights optimized
configuration for bf16, execution always happens in FP32 precision even
user manually set infer_precision=bf16.
This bug is not visible on FP16 IRs (since BF16+FP16 config is present),
so only FP32 IRs affected. SInce save_model and ovc apply FP16
compression be default, the issue mostly applicable for pipelines which
use a model directly after convert_model call.

Cherry-picks: https://github.com/openvinotoolkit/openvino/pull/22994
 		2024-02-22 11:31:35 +00:00
..
auto Updated copyright year to 2024 (#22958) 2024-02-20 17:01:21 +00:00
auto_batch Updated copyright year to 2024 (#22958) 2024-02-20 17:01:21 +00:00
hetero Updated copyright year to 2024 (#22958) 2024-02-20 17:01:21 +00:00
intel_cpu [CPU] Fixed BF16 Matmul inference precision (#22995) 2024-02-22 11:31:35 +00:00
intel_gpu Updated copyright year to 2024 (#22958) 2024-02-20 17:01:21 +00:00
proxy Updated copyright year to 2024 (#22958) 2024-02-20 17:01:21 +00:00
template Updated copyright year to 2024 (#22958) 2024-02-20 17:01:21 +00:00
CMakeLists.txt Updated copyright year to 2024 (#22958) 2024-02-20 17:01:21 +00:00
README.md [DOCS] Move of developer documentation from wiki to md documents - master (#15372) 2023-02-16 11:03:11 +04:00

README.md

OpenVINO Plugins

OpenVINO Plugins provide support for hardware devices.

The list of supported plugins:

See also