under test, not sure no errors
This commit is contained in:
30
upstream_ref/xllm/tools/README.md
Normal file
30
upstream_ref/xllm/tools/README.md
Normal file
@@ -0,0 +1,30 @@
|
||||
# NPU Timeline Generation Guide
|
||||
## Prerequisites
|
||||
- Python environment
|
||||
- Chrome browser (for visualization)
|
||||
## Implementation Steps
|
||||
### 1. Code Modification
|
||||
#### Register the subscriber
|
||||
Add the following at the beginning of your program:
|
||||
```cpp
|
||||
MsptiMetrics::register_subscriber();
|
||||
```
|
||||
#### Add tracing to ACLNN functions (work for msprof as well)
|
||||
Insert the following macro in your ACLNN functions where you want to measure performance:
|
||||
```cpp
|
||||
LLM_MSTX_RANGE();
|
||||
```
|
||||
#### Release the subscriber
|
||||
Add this at the end of your program:
|
||||
```cpp
|
||||
MsptiMetrics::release_subscriber();
|
||||
```
|
||||
### 2. Log Processing
|
||||
After running your program, process the generated log file using the timeline script:
|
||||
```bash
|
||||
python npu_timeline.py -i custom_log.log -o custom_output.json
|
||||
```
|
||||
### 3. Visualization
|
||||
Open Chrome browser
|
||||
Navigate to: chrome://tracing
|
||||
Load the generated JSON file: custom_output.json
|
||||
Reference in New Issue
Block a user