Artem Plastinkin commited on
Commit ·
187eea1
1
Parent(s): 91a1f75
Fix ONNX model name
Browse files
README.md
CHANGED
|
@@ -30,7 +30,7 @@ The FP32 ONNX model is auto-cast to **INT8** by the Renesas MWMX toolchain at co
|
|
| 30 |
separate quantization step is required.
|
| 31 |
|
| 32 |
```
|
| 33 |
-
|
| 34 |
│
|
| 35 |
└─▶ MWMX Runtime ──▶ INT8 auto-cast ──▶ NPX6 NPU
|
| 36 |
```
|
|
@@ -39,7 +39,7 @@ hrnet_w48_in1k.onnx (FP32)
|
|
| 39 |
|
| 40 |
| Artifact | Status | Notes |
|
| 41 |
|----------|--------|-------|
|
| 42 |
-
| **FP32 (ONNX)** | ⏳ Pending | `fp32/
|
| 43 |
|
| 44 |
## Performance
|
| 45 |
|
|
|
|
| 30 |
separate quantization step is required.
|
| 31 |
|
| 32 |
```
|
| 33 |
+
hrnet_w48_3rdparty_8xb32_in1k.onnx (FP32)
|
| 34 |
│
|
| 35 |
└─▶ MWMX Runtime ──▶ INT8 auto-cast ──▶ NPX6 NPU
|
| 36 |
```
|
|
|
|
| 39 |
|
| 40 |
| Artifact | Status | Notes |
|
| 41 |
|----------|--------|-------|
|
| 42 |
+
| **FP32 (ONNX)** | ⏳ Pending | `fp32/hrnet_w48_3rdparty_8xb32_in1k.onnx` — to be added; will be auto-cast to INT8 by the MWMX toolchain at compile time (see Deployment Flow above); no separate INT8 file will be shipped |
|
| 43 |
|
| 44 |
## Performance
|
| 45 |
|
int8/benchmarks/x5h_mwmx_npu_apm80_12core.yaml
CHANGED
|
@@ -53,7 +53,7 @@ reproduce:
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
-
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
|
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
+
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/hrnet_w48_3rdparty_8xb32_in1k.onnx --num-core 12
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
int8/benchmarks/x5h_mwmx_npu_apm80_1core.yaml
CHANGED
|
@@ -53,7 +53,7 @@ reproduce:
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
-
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
|
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
+
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/hrnet_w48_3rdparty_8xb32_in1k.onnx --num-core 1
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
int8/benchmarks/x5h_mwmx_npu_apm80_3core.yaml
CHANGED
|
@@ -53,7 +53,7 @@ reproduce:
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
-
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
|
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
+
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/hrnet_w48_3rdparty_8xb32_in1k.onnx --num-core 3
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
int8/benchmarks/x5h_mwmx_npu_apm80_4core.yaml
CHANGED
|
@@ -53,7 +53,7 @@ reproduce:
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
-
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
|
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
+
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/hrnet_w48_3rdparty_8xb32_in1k.onnx --num-core 4
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
int8/benchmarks/x5h_mwmx_npu_apm80_6core.yaml
CHANGED
|
@@ -53,7 +53,7 @@ reproduce:
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
-
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|
|
|
|
| 53 |
command: hf download Renesas/HRNet-ONNX --repo-type=model --include "fp32/*" --local-dir ./HRNet-ONNX-fp32
|
| 54 |
- title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
|
| 55 |
command: |
|
| 56 |
+
python3 nnac_frontend/legalize.py -d binary/nnx ./HRNet-ONNX-fp32/fp32/hrnet_w48_3rdparty_8xb32_in1k.onnx --num-core 6
|
| 57 |
- title: Set up the R-Car X5H board
|
| 58 |
command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
|
| 59 |
kind: note
|