Update at 2024-06-05 00:00:11.094250

mlc-ai · Jun 5, 2024 · aa9aa4b · aa9aa4b
1 parent cf4a625
commit aa9aa4b
Show file tree

Hide file tree

Showing 32 changed files with 2,152 additions and 1,312 deletions.
diff --git a/_images/output_index_1f4d27_59_0.png b/_images/output_index_1f4d27_59_0.png
diff --git a/_images/output_index_4f28a7_60_0.png b/_images/output_index_4f28a7_60_0.png
diff --git a/_images/output_index_e26dde_40_0.png b/_images/output_index_e26dde_40_0.png
diff --git a/_images/output_index_e758e2_5_0.png b/_images/output_index_e758e2_5_0.png
diff --git a/_sources/chapter_auto_program_optimization/index.rst.txt b/_sources/chapter_auto_program_optimization/index.rst.txt
diff --git a/_sources/chapter_end_to_end/index.rst.txt b/_sources/chapter_end_to_end/index.rst.txt
@@ -124,7 +124,7 @@ We can plot out the image instance that we want to be able to predict.
 .. parsed-literal::
     :class: output
 
-    Class: Sandal
+    Class: T-shirt/top
 
 
 Download Model Parameters
@@ -191,10 +191,10 @@ Let us begin by reviewing a Numpy implementation of the model.
 .. parsed-literal::
     :class: output
 
-    [[-26.194868  -34.8678    -24.12358   -20.510454  -18.109476   12.921284
-      -17.477413   -4.9815836  -7.2966995  -6.3170166]]
-    [5]
-    NumPy-MLP Prediction: Sandal
+    [[ 16.492514   -9.319946   -1.3977761  -7.6737294 -19.890661  -19.210732
+        7.6732693 -26.277058  -10.610296  -29.9112   ]]
+    [0]
+    NumPy-MLP Prediction: T-shirt/top
 
 
 The above example code shows the high-level array operations to perform
@@ -282,7 +282,7 @@ model.
 .. parsed-literal::
     :class: output
 
-    Low-level Numpy MLP Prediction: Sandal
+    Low-level Numpy MLP Prediction: T-shirt/top
 
 
 Constructing an End to End IRModule in TVMScript
@@ -528,7 +528,7 @@ the low-level numpy execution code as:
 .. parsed-literal::
     :class: output
 
-    Low-level Numpy with CallTIR Prediction: Sandal
+    Low-level Numpy with CallTIR Prediction: T-shirt/top
 
 
 In practice, the lowest-level implementation will have explicit memory
@@ -646,6 +646,7 @@ have.
     .output_html .cs { color: #3D7B7B; font-style: italic } /* Comment.Special */
     .output_html .gd { color: #A00000 } /* Generic.Deleted */
     .output_html .ge { font-style: italic } /* Generic.Emph */
+    .output_html .ges { font-weight: bold; font-style: italic } /* Generic.EmphStrong */
     .output_html .gr { color: #E40000 } /* Generic.Error */
     .output_html .gh { color: #000080; font-weight: bold } /* Generic.Heading */
     .output_html .gi { color: #008400 } /* Generic.Inserted */
@@ -837,8 +838,8 @@ weights.
 .. parsed-literal::
     :class: output
 
-    [[-26.19487   -34.867805  -24.123579  -20.510454  -18.109478   12.921284
-      -17.47741    -4.981583   -7.2966967  -6.3170156]]
+    [[ 16.49251    -9.319946   -1.3977766  -7.673729  -19.890663  -19.210732
+        7.67327   -26.277052  -10.610297  -29.911194 ]]
 
 
 The main function returns the prediction result, and we can then call
@@ -862,7 +863,7 @@ the class label.
 .. parsed-literal::
     :class: output
 
-    MyModule Prediction: Sandal
+    MyModule Prediction: T-shirt/top
 
 
 Integrate Existing Libraries in the Environment
@@ -993,7 +994,7 @@ that we get the same result.
 .. parsed-literal::
     :class: output
 
-    MyModuleWithExternCall Prediction: Sandal
+    MyModuleWithExternCall Prediction: T-shirt/top
 
 
 Mixing TensorIR Code and Libraries
@@ -1077,7 +1078,7 @@ functions. We can build and run to validate the result.
 .. parsed-literal::
     :class: output
 
-    MyModuleMixture Prediction: Sandal
+    MyModuleMixture Prediction: T-shirt/top
 
 
 Bind Parameters to IRModule
@@ -1122,6 +1123,7 @@ nd_params.
     .output_html .cs { color: #3D7B7B; font-style: italic } /* Comment.Special */
     .output_html .gd { color: #A00000 } /* Generic.Deleted */
     .output_html .ge { font-style: italic } /* Generic.Emph */
+    .output_html .ges { font-weight: bold; font-style: italic } /* Generic.EmphStrong */
     .output_html .gr { color: #E40000 } /* Generic.Error */
     .output_html .gh { color: #000080; font-weight: bold } /* Generic.Heading */
     .output_html .gi { color: #008400 } /* Generic.Inserted */
@@ -1250,7 +1252,7 @@ the input data.
 .. parsed-literal::
     :class: output
 
-    MyModuleWithParams Prediction: Sandal
+    MyModuleWithParams Prediction: T-shirt/top
 
 
 Discussions

diff --git a/_sources/chapter_gpu_acceleration/part1.rst.txt b/_sources/chapter_gpu_acceleration/part1.rst.txt
diff --git a/_sources/chapter_gpu_acceleration/part2.rst.txt b/_sources/chapter_gpu_acceleration/part2.rst.txt
diff --git a/_sources/chapter_graph_optimization/index.rst.txt b/_sources/chapter_graph_optimization/index.rst.txt
diff --git a/_sources/chapter_integration/index.rst.txt b/_sources/chapter_integration/index.rst.txt
diff --git a/_sources/chapter_tensor_program/case_study.rst.txt b/_sources/chapter_tensor_program/case_study.rst.txt
diff --git a/_sources/chapter_tensor_program/tensorir_exercises.rst.txt b/_sources/chapter_tensor_program/tensorir_exercises.rst.txt
@@ -69,9 +69,9 @@ computation abstraction (e.g., ``ndarray + ndarray``) to low-level
 python implementation (standard for loops with element access and
 operation)
 
-Notably, the initial value of the o utput array (or buffer) is not
-always ``0``. We need to write or initialize it in our implementation,
-which is important for reduction operator (e.g. matmul and conv)
+Notably, the initial value of the output array (or buffer) is not always
+``0``. We need to write or initialize it in our implementation, which is
+important for reduction operator (e.g. matmul and conv)
 
 .. raw:: latex
 
@@ -315,8 +315,8 @@ Parallel, Vectorize and Unroll
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 First, we introduce some new primitives, ``parallel``, ``vectorize`` and
-``unroll``. These three primitives operates on loops to indicate how
-this loop execute. Here is the example:
+``unroll``. These three primitives operate on loops to indicate how this
+loop executes. Here is the example:
 
 .. raw:: latex
 
@@ -370,6 +370,7 @@ this loop execute. Here is the example:
     .output_html .cs { color: #3D7B7B; font-style: italic } /* Comment.Special */
     .output_html .gd { color: #A00000 } /* Generic.Deleted */
     .output_html .ge { font-style: italic } /* Generic.Emph */
+    .output_html .ges { font-weight: bold; font-style: italic } /* Generic.EmphStrong */
     .output_html .gr { color: #E40000 } /* Generic.Error */
     .output_html .gh { color: #000080; font-weight: bold } /* Generic.Heading */
     .output_html .gi { color: #008400 } /* Generic.Inserted */
@@ -431,7 +432,6 @@ this loop execute. Here is the example:
     <span class="k">class</span> <span class="nc">Module</span><span class="p">:</span>
         <span class="nd">@T</span><span class="o">.</span><span class="n">prim_func</span>
         <span class="k">def</span> <span class="nf">add</span><span class="p">(</span><span class="n">A</span><span class="p">:</span> <span class="n">T</span><span class="o">.</span><span class="n">Buffer</span><span class="p">((</span><span class="mi">4</span><span class="p">,</span> <span class="mi">4</span><span class="p">),</span> <span class="s2">&quot;int64&quot;</span><span class="p">),</span> <span class="n">B</span><span class="p">:</span> <span class="n">T</span><span class="o">.</span><span class="n">Buffer</span><span class="p">((</span><span class="mi">4</span><span class="p">,</span> <span class="mi">4</span><span class="p">),</span> <span class="s2">&quot;int64&quot;</span><span class="p">),</span> <span class="n">C</span><span class="p">:</span> <span class="n">T</span><span class="o">.</span><span class="n">Buffer</span><span class="p">((</span><span class="mi">4</span><span class="p">,</span> <span class="mi">4</span><span class="p">),</span> <span class="s2">&quot;int64&quot;</span><span class="p">)):</span>
-            <span class="n">T</span><span class="o">.</span><span class="n">func_attr</span><span class="p">({</span><span class="s2">&quot;global_symbol&quot;</span><span class="p">:</span> <span class="s2">&quot;add&quot;</span><span class="p">})</span>
             <span class="c1"># with T.block(&quot;root&quot;):</span>
             <span class="k">for</span> <span class="n">i_0</span> <span class="ow">in</span> <span class="n">T</span><span class="o">.</span><span class="n">parallel</span><span class="p">(</span><span class="mi">2</span><span class="p">):</span>
                 <span class="k">for</span> <span class="n">i_1</span> <span class="ow">in</span> <span class="n">T</span><span class="o">.</span><span class="n">unroll</span><span class="p">(</span><span class="mi">2</span><span class="p">):</span>

diff --git a/_static/mlc-favicon.ico b/_static/mlc-favicon.ico
diff --git a/_static/mlc-favicon.png b/_static/mlc-favicon.png
diff --git a/_static/pygments.css b/_static/pygments.css
@@ -17,6 +17,7 @@ span.linenos.special { color: #000000; background-color: #ffffc0; padding-left:
 .highlight .cs { color: #60a0b0; background-color: #fff0f0 } /* Comment.Special */
 .highlight .gd { color: #A00000 } /* Generic.Deleted */
 .highlight .ge { font-style: italic } /* Generic.Emph */
+.highlight .ges { font-weight: bold; font-style: italic } /* Generic.EmphStrong */
 .highlight .gr { color: #FF0000 } /* Generic.Error */
 .highlight .gh { color: #000080; font-weight: bold } /* Generic.Heading */
 .highlight .gi { color: #00A000 } /* Generic.Inserted */