Add CI for tt-explorer (#1109)

tenstorrent · Nov 1, 2024 · c78b606 · c78b606
1 parent 8e939d9
commit c78b606
Show file tree

Hide file tree

Showing 3 changed files with 68 additions and 2 deletions.
diff --git a/.github/workflows/build-and-test.yml b/.github/workflows/build-and-test.yml
@@ -404,3 +404,69 @@ jobs:
       run: |
         source env/activate
         pytest -ssv runtime/tools/python/test/test_perf.py
+
+  build-and-test-explorer:
+    needs: build-image
+    timeout-minutes: 60
+    strategy:
+      fail-fast: false
+      matrix:
+        build: [
+          {runs-on: ubuntu-latest, enable_perf: OFF, name: "run", ttrt_flags: ""},
+        ]
+
+    name: Build and test tt-explorer
+    runs-on: ${{ matrix.build.runs-on }}
+
+    container:
+      image: ${{ needs.build-image.outputs.docker-image }}
+
+    steps:
+
+    - uses: actions/checkout@v4
+      with:
+          fetch-depth: 0
+
+    - name: Set reusable strings
+      id: strings
+      shell: bash
+      run: |
+        echo "work-dir=$(pwd)" >> "$GITHUB_OUTPUT"
+        echo "build-output-dir=$(pwd)/build" >> "$GITHUB_OUTPUT"
+
+    - name: Git safe dir
+      run: git config --global --add safe.directory ${{ steps.strings.outputs.work-dir }}
+
+    - name: ccache
+      uses: hendrikmuhs/ccache-action@v1.2
+      with:
+        create-symlink: true
+        key: ${{ matrix.build.runs-on }}-run-ON-perf-${{ matrix.build.enable_perf }}-${{ env.SDK_VERSION }}
+
+    - name: Configure CMake
+      shell: bash
+      run: |
+        source env/activate
+        cmake -G Ninja \
+        -B ${{ steps.strings.outputs.build-output-dir }} \
+        -DCMAKE_CXX_COMPILER=clang++-17 \
+        -DCMAKE_C_COMPILER=clang-17 \
+        -DCMAKE_BUILD_TYPE=Release \
+        -DCMAKE_CXX_COMPILER_LAUNCHER=ccache \
+        -DTTMLIR_ENABLE_RUNTIME=OFF \
+        -DTTMLIR_ENABLE_RUNTIME_TESTS=OFF \
+        -DTT_RUNTIME_ENABLE_PERF_TRACE=${{ matrix.build.enable_perf }} \
+        -DTTMLIR_ENABLE_STABLEHLO=OFF \
+        -S ${{ steps.strings.outputs.work-dir }}
+
+    - name: Build tt-explorer
+      shell: bash
+      run: |
+        source env/activate
+        cmake --build ${{ steps.strings.outputs.build-output-dir }} -- explorer
+
+    - name: Run tt-explorer tests
+      shell: bash
+      run: |
+        source env/activate
+        pytest tools/explorer/test/run_tests.py
diff --git a/tools/explorer/test/models/forward_and_backward.mlir b/tools/explorer/test/models/forward_and_backward.mlir
@@ -1,4 +1,4 @@
-module @SimpleModel attributes {tt.system_desc = #tt.system_desc<[{arch = <wormhole_b0>, grid = 8x8, l1_size = 1499136, num_dram_channels = 12, dram_channel_size = 1073741824, noc_l1_address_align_bytes = 16, pcie_address_align_bytes = 32, noc_dram_address_align_bytes = 32, l1_unreserved_base = 1024, erisc_l1_unreserved_base = 1024, dram_unreserved_base = 1024, dram_unreserved_end = 1073741824, physical_cores = {worker = [ 0x0,  0x1,  0x2,  0x3,  0x4,  0x5,  0x6,  0x7,  1x0,  1x1,  1x2,  1x3,  1x4,  1x5,  1x6,  1x7,  2x0,  2x1,  2x2,  2x3,  2x4,  2x5,  2x6,  2x7,  3x0,  3x1,  3x2,  3x3,  3x4,  3x5,  3x6,  3x7,  4x0,  4x1,  4x2,  4x3,  4x4,  4x5,  4x6,  4x7,  5x0,  5x1,  5x2,  5x3,  5x4,  5x5,  5x6,  5x7,  6x0,  6x1,  6x2,  6x3,  6x4,  6x5,  6x6,  6x7,  7x0,  7x1,  7x2,  7x3,  7x4,  7x5,  7x6,  7x7] dram = [ 8x0,  9x0,  10x0,  8x1,  9x1,  10x1,  8x2,  9x2,  10x2,  8x3,  9x3,  10x3]}, supported_data_types = [<f32>, <f16>, <bf16>, <bfp_f8>, <bfp_bf8>, <bfp_f4>, <bfp_bf4>, <bfp_f2>, <bfp_bf2>, <u32>, <u16>, <u8>], supported_tile_sizes = [ 4x16,  16x16,  32x16,  4x32,  16x32,  32x32]}], [0], [3 : i32], [ 0x0x0x0]>} {
+module @SimpleModel attributes {} {
   func.func @forward(%arg0: tensor<1x784xf32> {ttir.name = "input_1"}, %arg1: tensor<10x784xf32> {ttir.name = "linear.weight"}, %arg2: tensor<10xf32> {ttir.name = "linear.bias"}) -> (tensor<1x10xf32> {ttir.name = "SimpleModel_472.output_softmax_1495"}) {
     %0 = tensor.empty() : tensor<784x10xf32>
     %1 = "ttir.transpose"(%arg1, %0) <{dim0 = -2 : si32, dim1 = -1 : si32, operand_constraints = [#tt.operand_constraint<dram|l1|scalar|tile|none|interleaved|single_bank|height_sharded|width_sharded|block_sharded|any_layout|any_device|any_device_tile|l1_block_sharded>, #tt.operand_constraint<dram|l1|scalar|tile|none|interleaved|single_bank|height_sharded|width_sharded|block_sharded|any_layout|any_device|any_device_tile|l1_block_sharded>]}> : (tensor<10x784xf32>, tensor<784x10xf32>) -> tensor<784x10xf32>

diff --git a/tools/explorer/test/models/linear_autoencoder.mlir b/tools/explorer/test/models/linear_autoencoder.mlir
@@ -1,4 +1,4 @@
-module @LinearAE attributes {tt.system_desc = #tt.system_desc<[{arch = <wormhole_b0>, grid = 8x8, l1_size = 1499136, num_dram_channels = 12, dram_channel_size = 1073741824, noc_l1_address_align_bytes = 16, pcie_address_align_bytes = 32, noc_dram_address_align_bytes = 32, l1_unreserved_base = 1024, erisc_l1_unreserved_base = 1024, dram_unreserved_base = 1024, dram_unreserved_end = 1073741824, physical_cores = {worker = [ 0x0,  0x1,  0x2,  0x3,  0x4,  0x5,  0x6,  0x7,  1x0,  1x1,  1x2,  1x3,  1x4,  1x5,  1x6,  1x7,  2x0,  2x1,  2x2,  2x3,  2x4,  2x5,  2x6,  2x7,  3x0,  3x1,  3x2,  3x3,  3x4,  3x5,  3x6,  3x7,  4x0,  4x1,  4x2,  4x3,  4x4,  4x5,  4x6,  4x7,  5x0,  5x1,  5x2,  5x3,  5x4,  5x5,  5x6,  5x7,  6x0,  6x1,  6x2,  6x3,  6x4,  6x5,  6x6,  6x7,  7x0,  7x1,  7x2,  7x3,  7x4,  7x5,  7x6,  7x7] dram = [ 8x0,  9x0,  10x0,  8x1,  9x1,  10x1,  8x2,  9x2,  10x2,  8x3,  9x3,  10x3]}, supported_data_types = [<f32>, <f16>, <bf16>, <bfp_f8>, <bfp_bf8>, <bfp_f4>, <bfp_bf4>, <bfp_f2>, <bfp_bf2>, <u32>, <u16>, <u8>], supported_tile_sizes = [ 4x16,  16x16,  32x16,  4x32,  16x32,  32x32]}], [0], [3 : i32], [ 0x0x0x0]>} {
+module @LinearAE attributes {} {
   func.func @forward(%arg0: tensor<1x784xf32> {ttir.name = "input_1"}, %arg1: tensor<784x128xf32> {ttir.name = "encoder_lin1.weight"}, %arg2: tensor<128xf32> {ttir.name = "encoder_lin1.bias"}, %arg3: tensor<128x64xf32> {ttir.name = "encoder_lin2.weight"}, %arg4: tensor<64xf32> {ttir.name = "encoder_lin2.bias"}, %arg5: tensor<64x12xf32> {ttir.name = "encoder_lin3.weight"}, %arg6: tensor<12xf32> {ttir.name = "encoder_lin3.bias"}, %arg7: tensor<12x3xf32> {ttir.name = "encoder_lin4.weight"}, %arg8: tensor<3xf32> {ttir.name = "encoder_lin4.bias"}, %arg9: tensor<3x12xf32> {ttir.name = "decoder_lin1.weight"}, %arg10: tensor<12xf32> {ttir.name = "decoder_lin1.bias"}, %arg11: tensor<12x64xf32> {ttir.name = "decoder_lin2.weight"}, %arg12: tensor<64xf32> {ttir.name = "decoder_lin2.bias"}, %arg13: tensor<64x128xf32> {ttir.name = "decoder_lin3.weight"}, %arg14: tensor<128xf32> {ttir.name = "decoder_lin3.bias"}, %arg15: tensor<128x784xf32> {ttir.name = "decoder_lin4.weight"}, %arg16: tensor<784xf32> {ttir.name = "decoder_lin4.bias"}) -> (tensor<1x784xf32> {ttir.name = "LinearAE.output_add_29"}) {
     %0 = tensor.empty() : tensor<1x128xf32>
     %1 = "ttir.matmul"(%arg0, %arg1, %0) <{operand_constraints = [#tt.operand_constraint<dram|l1|scalar|tile|none|interleaved|single_bank|height_sharded|width_sharded|block_sharded|any_layout|any_device|any_device_tile|l1_block_sharded>, #tt.operand_constraint<dram|l1|scalar|tile|none|interleaved|single_bank|height_sharded|width_sharded|block_sharded|any_layout|any_device|any_device_tile|l1_block_sharded>, #tt.operand_constraint<dram|l1|scalar|tile|none|interleaved|single_bank|height_sharded|width_sharded|block_sharded|any_layout|any_device|any_device_tile|l1_block_sharded>]}> : (tensor<1x784xf32>, tensor<784x128xf32>, tensor<1x128xf32>) -> tensor<1x128xf32>