Update README.md

README.md

@@ -1,6 +1,4 @@
 ---
-base_model:
-- openbmb/MiniCPM-V-2_6
 tags:
 - rknn
 - rkllm
@@ -39,63 +37,46 @@ python multiprocess_inference.py
 
 If the performance is not satisfactory, you can change the CPU scheduler to keep the CPU running at the highest frequency, and bind the inference program to the big core cluster (`taskset -c 4-7 python multiprocess_inference.py`).
 
-test.jpg:
+man.jpg:
 ![man.jpg](./man.jpg)
 
->```
->Start loading language model (size: 7810.02 MB)
->
->I rkllm: rkllm-runtime version: 1.1.2, rknpu driver version: 0.9.8, platform: RK3588
->
->W rknn-toolkit-lite2 version: 2.2.0
->Start loading vision encoder model (size: 942.29 MB)
->Vision encoder loaded in 10.22 seconds
->I RKNN: [02:28:20.939] RKNN Runtime Information, librknnrt version: 2.1.0 (967d001cc8@2024-08-07T19:28:19)
->I RKNN: [02:28:20.939] RKNN Driver Information, version: 0.9.8
->I RKNN: [02:28:20.940] RKNN Model Information, version: 6, toolkit version: 2.2.0(compiler version: 2.2.0 (c195366594@2024-09-14T12:24:14)), target: RKNPU v2, target platform: rk3588, framework name: ONNX, framework layout: NCHW, model inference type: dynamic_shape
->W RKNN: [02:28:20.940] RKNN Model version: 2.2.0 not match with rknn runtime version: 2.1.0
->Received ready signal: vision_ready
->Language model loaded in 29.21 seconds
->Received ready signal: llm_ready
->All models loaded, starting interactive mode...
->
->Enter your input (3 empty lines to start inference, Ctrl+C to exit, for example: 
->详细描述一下{{./test.jpg}}这张图片
->What is the weather in {{./test.jpg}}?
->How many people are in {{./test.jpg}}?
->):
->
->Describe the image: {{test.jpg}} in every detail.
->
->
->
->Start vision inference...
->
->Vision encoder inference time: 3.26 seconds
->
->Time to first token: 1.72 seconds
->
->The image depicts an urban street scene with various elements that contribute to its bustling atmosphere.
->
->A person, likely male based on appearance, is walking across the crosswalk carrying a blue and white checked umbrella. He's dressed casually yet stylishly, wearing a beige jacket over what appears to be dark pants or leggings paired with patterned slip-on shoes in shades of gray, black, and yellow.
->
->The street itself features multiple lanes filled with vehicles; there are cars visible on both sides, including a prominent SUV that is parked by the roadside. The presence of these automobiles adds to the sense of movement and activity within this urban setting.
->
->In terms of infrastructure, the crosswalk has clear pedestrian markings for safety, and an adjacent railing provides support or boundary along one side of the street. Beyond the immediate foreground where pedestrians traverse, there's a sidewalk lined with lush green trees which add natural beauty to the otherwise concrete-dominated environment.
->
->The sky is visible in parts through breaks in clouds above, indicating fair weather conditions that contribute positively to outdoor activities like walking down this cityscape path.
->
->Overall, it appears as though an ordinary day unfolds within this urban setting, capturing moments of daily life and movement.
->
->(finished)
->
->--------------------------------------------------------------------------------------
-> Stage         Total Time (ms)  Tokens    Time per Token (ms)      Tokens per Second       
->--------------------------------------------------------------------------------------
-> Prefill       1714.78          94        18.24                    54.82                   
-> Generate      58689.71         236       249.75                   4.00                    
->--------------------------------------------------------------------------------------
->```
+```bash
+admin@orangepi5:~/MiniCPM-V-2_6-rkllm$ python multiprocess_inference.py
+Start loading language model (size: 7810.02 MB)
+I rkllm: rkllm-runtime version: 1.1.4, rknpu driver version: 0.9.8, platform: RK3588
+
+W rknn-toolkit-lite2 version: 2.3.0
+Start loading vision encoder model (size: 942.29 MB)
+Vision encoder loaded in 4.95 seconds
+I RKNN: [13:13:11.477] RKNN Runtime Information, librknnrt version: 2.3.0 (c949ad889d@2024-11-07T11:35:33)
+I RKNN: [13:13:11.477] RKNN Driver Information, version: 0.9.8
+I RKNN: [13:13:11.478] RKNN Model Information, version: 6, toolkit version: 2.2.0(compiler version: 2.2.0 (c195366594@2024-09-14T12:24:14)), target: RKNPU v2, target platform: rk3588, framework name: ONNX, framework layout: NCHW, model inference type: dynamic_shape
+Received ready signal: vision_ready
+Language model loaded in 30.56 seconds
+Received ready signal: llm_ready
+All models loaded, starting interactive mode...
+
+Enter your input :
+
+Describe the person in the image {{./man.jpg}}?
+How many people are in he image {{./man.jpg}}?
+
+
+
+Start vision inference...
+Vision encoder inference time: 3.61 seconds
+Time to first token: 2.09 seconds
+There is one person in the image the second image . The focus of this photograph seems to be on an individual, as there are no other discernible figures present. Please note that I can only provide descriptions based on what's visible within the images you've shared. If you have any specific questions about the content or need further details, feel free to ask!
+
+(finished)
+
+--------------------------------------------------------------------------------------
+ Stage         Total Time (ms)  Tokens    Time per Token (ms)      Tokens per Second
+--------------------------------------------------------------------------------------
+ Prefill       2040.37          112       18.22                    54.89
+ Generate      22301.36         72        310.92                   3.22
+--------------------------------------------------------------------------------------
+```
 
 ## Model Conversion