Xidong commited on
Commit
077252c
·
verified ·
1 Parent(s): 3c9587c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +84 -0
README.md CHANGED
@@ -1,3 +1,87 @@
1
  ---
2
  license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+ # EmbodyAICar
5
+ Embody_AI with car as Demo
6
+
7
+ <center>
8
+ <img src="https://github.com/FreedomIntelligence/EmbodyAICar/blob/main/assets/embodyaicar.png" width="450" style="display: inline-block; vertical-align: top;"/>
9
+ </center>
10
+
11
+
12
+
13
+
14
+ <p align="center">
15
+ 🌐 <a href="" target="_blank">Website</a> • 🤗 <a href="" target="_blank">Model</a>
16
+ </p>
17
+
18
+ ## 🌈 Update
19
+ * **[2024.02.23]** 🎉🎉🎉 EmbodyAICar is published!🎉🎉🎉
20
+
21
+ ## 🤖 Model Training Data
22
+
23
+ ```
24
+ User:快速向左转
25
+ RobotAI: (1.0, -0.3)
26
+ ...
27
+ ```
28
+
29
+ - The two float are in range [-1,1]
30
+ - The first float is speed, the second is direction (negative means left, positive means right).
31
+
32
+ ## 🤖 Communication Protocol
33
+
34
+ - 0x02, 0x02, 0x01, 8, data_buf; (See detail in [code](https://github.com/FreedomIntelligence/EmbodyAICar/blob/main/qwen.cpp#L151))
35
+
36
+
37
+ ## ℹ️ Usage
38
+ 1. DownLoad [Model]() and Follow [Qwen.cpp](https://github.com/QwenLM/qwen.cpp.git) get model.bin and qwen.tiktoken.
39
+ 2. Install package serial.tar.gz
40
+ ```
41
+ cd serial
42
+ cmake .. & make & sudo make install
43
+ ```
44
+ 3. Compile the project using CMake:
45
+ ```
46
+ cmake -B build
47
+ cmake --build build -j --config Release
48
+ ```
49
+ 4. Now you may chat and control your AI car with the quantized RobotAI model by running:
50
+ ```
51
+ ./build/bin/main -m robot1_8b-ggml.bin --tiktoken qwen.tiktoken -p 请快速向前
52
+ ```
53
+
54
+ To run the model in interactive mode, add the -i flag. For example:
55
+
56
+ ```
57
+ ./build/bin/main -m robot1_8b-ggml.bin --tiktoken qwen.tiktoken -i
58
+ ```
59
+
60
+ In interactive mode, your chat history will serve as the context for the next-round conversation.
61
+
62
+ ## 🥸 To do list
63
+
64
+ - Continue to create data and train a robust model
65
+ - Add ASR and TTS
66
+ - ...
67
+
68
+
69
+ ## ✨ Citation
70
+ Please use the following citation if you intend to use our dataset for training or evaluation:
71
+
72
+
73
+
74
+ ```
75
+ @misc{embodyAiCar,
76
+ title={EmbodyAICar: Embody_AI with car as Demo},
77
+ author={Xidong Wang*, Yuan Shen*},
78
+ year = {2024},
79
+ publisher = {GitHub},
80
+ journal = {GitHub repository},
81
+ howpublished = {\url{https://github.com/FreedomIntelligence/EmbodyAICar}},
82
+ }
83
+ ```
84
+
85
+
86
+ ## 🤖 Acknowledgement
87
+ - We thank [Qwen.cpp](https://github.com/QwenLM/qwen.cpp.git) and [llama.cpp](https://github.com/ggerganov/llama.cpp) for their excellent work.