Skip to content

Commit 1824cb2

Browse files
author
TRTorch Github Bot
committed
docs: [Automated] Regenerating documenation from
Signed-off-by: TRTorch Github Bot <[email protected]>
1 parent be080a9 commit 1824cb2

10 files changed

+39
-5
lines changed

‎docs/_cpp_api/program_listing_file_cpp_api_include_trtorch_trtorch.h.html

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -467,6 +467,8 @@ <h1 id="cpp-api-program-listing-file-cpp-api-include-trtorch-trtorch-h--page-roo
467467

468468
<span class="n">DataType</span> <span class="n">op_precision</span> <span class="o">=</span> <span class="n">DataType</span><span class="o">::</span><span class="n">kFloat</span><span class="p">;</span>
469469

470+
<span class="kt">bool</span> <span class="n">disable_tf32</span> <span class="o">=</span> <span class="nb">false</span><span class="p">;</span>
471+
470472
<span class="kt">bool</span> <span class="n">refit</span> <span class="o">=</span> <span class="nb">false</span><span class="p">;</span>
471473

472474
<span class="kt">bool</span> <span class="n">debug</span> <span class="o">=</span> <span class="nb">false</span><span class="p">;</span>

‎docs/_cpp_api/structtrtorch_1_1CompileSpec.html

Lines changed: 23 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -793,6 +793,29 @@ <h2 id="struct-documentation">
793793
</p>
794794
</dd>
795795
</dl>
796+
<dl class="cpp var">
797+
<dt id="_CPPv4N7trtorch11CompileSpec12disable_tf32E">
798+
<span class="target" id="structtrtorch_1_1CompileSpec_1a0daef036d4cbdac49dda3c9cd284a661">
799+
</span>
800+
bool
801+
<code class="sig-name descname">
802+
disable_tf32
803+
</code>
804+
= false
805+
<a class="headerlink" href="#_CPPv4N7trtorch11CompileSpec12disable_tf32E" title="Permalink to this definition">
806+
807+
</a>
808+
<br/>
809+
</dt>
810+
<dd>
811+
<p>
812+
Prevent Float32 layers from using TF32 data format
813+
</p>
814+
<p>
815+
TF32 computes inner products by rounding the inputs to 10-bit mantissas before multiplying, but accumulates the sum using 23-bit mantissas. This is the behavior of FP32 layers by default.
816+
</p>
817+
</dd>
818+
</dl>
796819
<dl class="cpp var">
797820
<dt id="_CPPv4N7trtorch11CompileSpec5refitE">
798821
<span class="target" id="structtrtorch_1_1CompileSpec_1a637f91d7d5b05ad01d7f4a38ab1e4f46">

‎docs/_notebooks/Resnet50-example.html

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -675,7 +675,7 @@
675675
</div>
676676
</div>
677677
<p>
678-
<img alt="e8cc1e76fbec4ae18bcf6bf7afd29fa5" src="http://developer.download.nvidia.com/compute/machine-learning/frameworks/nvidia_logo.png"/>
678+
<img alt="70511f56741d40949896fdabbc794e22" src="http://developer.download.nvidia.com/compute/machine-learning/frameworks/nvidia_logo.png"/>
679679
</p>
680680
<h1 id="notebooks-resnet50-example--page-root">
681681
TRTorch Getting Started - ResNet 50

‎docs/_notebooks/lenet-getting-started.html

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -769,7 +769,7 @@
769769
</div>
770770
</div>
771771
<p>
772-
<img alt="6bbcefe2c1b143878ad41f204a6698b5" src="http://developer.download.nvidia.com/compute/machine-learning/frameworks/nvidia_logo.png"/>
772+
<img alt="4aab7546fe7547e68f367db08f8f5138" src="http://developer.download.nvidia.com/compute/machine-learning/frameworks/nvidia_logo.png"/>
773773
</p>
774774
<h1 id="notebooks-lenet-getting-started--page-root">
775775
TRTorch Getting Started - LeNet

‎docs/_notebooks/ssd-object-detection-demo.html

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -789,7 +789,7 @@
789789
</div>
790790
</div>
791791
<p>
792-
<img alt="e0d693b92c2c4e03921a2b446e3e44d1" src="http://developer.download.nvidia.com/compute/machine-learning/frameworks/nvidia_logo.png"/>
792+
<img alt="75476f2f22d44824a3852a4296681786" src="http://developer.download.nvidia.com/compute/machine-learning/frameworks/nvidia_logo.png"/>
793793
</p>
794794
<h1 id="notebooks-ssd-object-detection-demo--page-root">
795795
Object Detection with TRTorch (SSD)

‎docs/_sources/_cpp_api/program_listing_file_cpp_api_include_trtorch_trtorch.h.rst.txt

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -107,6 +107,8 @@ Program Listing for File trtorch.h
107107
108108
DataType op_precision = DataType::kFloat;
109109
110+
bool disable_tf32 = false;
111+
110112
bool refit = false;
111113
112114
bool debug = false;

‎docs/genindex.html

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1038,6 +1038,11 @@ <h2 id="T">
10381038
[1]
10391039
</a>
10401040
</li>
1041+
<li>
1042+
<a href="_cpp_api/structtrtorch_1_1CompileSpec.html#_CPPv4N7trtorch11CompileSpec12disable_tf32E">
1043+
trtorch::CompileSpec::disable_tf32 (C++ member)
1044+
</a>
1045+
</li>
10411046
<li>
10421047
<a href="_cpp_api/structtrtorch_1_1CompileSpec.html#_CPPv4N7trtorch11CompileSpec16EngineCapabilityE">
10431048
trtorch::CompileSpec::EngineCapability (C++ enum)

‎docs/objects.inv

58 Bytes
Binary file not shown.

‎docs/py_api/trtorch.html

Lines changed: 3 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -721,6 +721,7 @@ <h2 id="functions">
721721
<span class="s2">"allow_gpu_fallback"</span><span class="p">:</span> <span class="n">false</span><span class="p">,</span> <span class="c1"># (DLA only) Allow layers unsupported on DLA to run on GPU</span>
722722
<span class="p">},</span>
723723
<span class="s2">"op_precision"</span><span class="p">:</span> <span class="n">torch</span><span class="o">.</span><span class="n">half</span><span class="p">,</span> <span class="c1"># Operating precision set to FP16</span>
724+
<span class="s2">"disable_tf32"</span><span class="p">:</span> <span class="kc">False</span><span class="p">,</span> <span class="c1"># Force FP32 layers to use traditional as FP32 format vs the default behavior of rounding the inputs to 10-bit mantissas before multiplying, but accumulates the sum using 23-bit mantissas</span>
724725
<span class="s2">"refit"</span><span class="p">:</span> <span class="n">false</span><span class="p">,</span> <span class="c1"># enable refit</span>
725726
<span class="s2">"debug"</span><span class="p">:</span> <span class="n">false</span><span class="p">,</span> <span class="c1"># enable debuggable engine</span>
726727
<span class="s2">"strict_types"</span><span class="p">:</span> <span class="n">false</span><span class="p">,</span> <span class="c1"># kernels should strictly run in operating precision</span>
@@ -969,7 +970,7 @@ <h2 id="functions">
969970
<span class="sig-paren">
970971
)
971972
</span>
972-
→ &lt;torch._C.ScriptClass object at 0x7f1a30a16a08&gt;
973+
→ &lt;torch._C.ScriptClass object at 0x7f927f22e228&gt;
973974
<a class="headerlink" href="#trtorch.TensorRTCompileSpec" title="Permalink to this definition">
974975
975976
</a>
@@ -1023,6 +1024,7 @@ <h2 id="functions">
10231024
<span class="s2">"allow_gpu_fallback"</span><span class="p">:</span> <span class="n">false</span><span class="p">,</span> <span class="c1"># (DLA only) Allow layers unsupported on DLA to run on GPU</span>
10241025
<span class="p">},</span>
10251026
<span class="s2">"op_precision"</span><span class="p">:</span> <span class="n">torch</span><span class="o">.</span><span class="n">half</span><span class="p">,</span> <span class="c1"># Operating precision set to FP16</span>
1027+
<span class="s2">"disable_tf32"</span><span class="p">:</span> <span class="kc">False</span><span class="p">,</span> <span class="c1"># Force FP32 layers to use traditional as FP32 format vs the default behavior of rounding the inputs to 10-bit mantissas before multiplying, but accumulates the sum using 23-bit mantissas</span>
10261028
<span class="s2">"refit"</span><span class="p">:</span> <span class="kc">False</span><span class="p">,</span> <span class="c1"># enable refit</span>
10271029
<span class="s2">"debug"</span><span class="p">:</span> <span class="kc">False</span><span class="p">,</span> <span class="c1"># enable debuggable engine</span>
10281030
<span class="s2">"strict_types"</span><span class="p">:</span> <span class="kc">False</span><span class="p">,</span> <span class="c1"># kernels should strictly run in operating precision</span>

‎docs/searchindex.js

Lines changed: 1 addition & 1 deletion
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

0 commit comments

Comments
 (0)