TinyML

来自Jack's Lab
（版本间的差异）
跳转到： 导航, 搜索
(→‎Quick Start)
(→‎Overview)
 
（未显示1个用户的11个中间版本）
第3行： 第3行：
 
* [https://eloquentarduino.github.io/2020/01/easy-tinyml-on-esp32-and-arduino/ Easy Tensorflow TinyML on ESP32 and Arduino]
 
* [https://eloquentarduino.github.io/2020/01/easy-tinyml-on-esp32-and-arduino/ Easy Tensorflow TinyML on ESP32 and Arduino]
 
* z-lib.org
 
* z-lib.org
 +
 +
 +
 +
TinyML 基金会在 2019 年组织了第一届峰会，这届峰会的成果如下：
 +
 +
* TinyML 的技术硬件已经进入了实用性的阶段；
 +
* 算法，网络以及低于 100KB 的 ML 模型，已经取得重大突破；
 +
* 视觉，音频的低功耗需求快速增长。
  
 
<br><br>
 
<br><br>
 +
 +
== Hardware ==
 +
 +
* Sparkfunc Edge:
 +
**  Cortex-M4F up to 48MHz, 6uA/MHz (Ambiq Micro Apollo3 Blue) https://ambiq.com/apollo3-blue-datasheet/
 +
** 1MB Flash, 384KB RAM
 +
 +
 +
* Arduino Nano 33 BLE Sense [https://docs.arduino.cc/static/5ae03c9567ffd89ec14cfc856a5d58a2/ABX00031-datasheet.pdf Nano 33 Datasheet]
 +
** Cortex-M4F, up to 64MHz, 52 µA/MHz (Nordic nRF52480) https://content.arduino.cc/assets/Nano_BLE_MCU-nRF52840_PS_v1.1.pdf
 +
** 1 MB Flash, 256 kB RAM
 +
 +
 +
* [https://www.st.com/en/evaluation-tools/32f746gdiscovery.html  STM32F746G Discovery Kit]
 +
** Cortex®-M7 up to 216MHz [https://www.st.com/resource/en/datasheet/stm32f746ng.pdf STM32F746G DS]
 +
** 1 MB Flash, 320KB RAM
 +
** LQFP100 (14x14 mm), TFBGA100 (8x8 mm), WLCSP143 (4.5x5.8 mm)
 +
 +
<br>
  
 
== Quick Start ==
 
== Quick Start ==
  
 
<source lang=bash>
 
<source lang=bash>
−
$ pip3 install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow-gpu==2.3.0
+
$ pip3 install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow-gpu==2.5.0
  
 
$ python3
 
$ python3
 
import tensorflow as tf
 
import tensorflow as tf
−
hello = tf.constant('Hello, TensorFlow!')
+
>>> print(tf.__version__)
−
sess = tf.Session()
+
2.5.0
−
print(sess.run(hello))
+
 
−
+
>>> print('GPU: ', tf.test.is_gpu_available())
−
Hello, TensorFlow!
+
2022-04-01 03:24:13.099644: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1733] Found device 0 with properties:
 +
pciBusID: 0000:01:00.0 name: NVIDIA GeForce GTX 1050 computeCapability: 6.1
 +
coreClock: 1.493GHz coreCount: 5 deviceMemorySize: 4.00GiB deviceMemoryBandwidth: 104.43GiB/s
 +
2022-04-01 03:24:13.099980: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1871] Adding visible gpu devices: 0
 +
2022-04-01 03:24:13.100248: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1258] Device interconnect StreamExecutor with strength 1 edge matrix:
 +
2022-04-01 03:24:13.100459: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1264]      0
 +
2022-04-01 03:24:13.100669: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1277] 0:  N
 +
2022-04-01 03:24:13.100967: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1418] Created TensorFlow device (/device:GPU:0 with 2779 MB memory) -> physical G
 +
PU (device: 0, name: NVIDIA GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1)
 +
GPU:  True
 +
 
 +
>>> hello = tf.constant('Hello, Tensorflow!')
 +
>>> print(hello)
 +
tf.Tensor(b'Hello, Tensorflow!', shape=(), dtype=string)
 +
 
 +
>>> from tensorflow.python.client import device_lib
 +
>>> print(device_lib.list_local_devices())
 +
[name: "/device:CPU:0"
 +
device_type: "CPU"
 +
memory_limit: 268435456
 +
locality {
 +
}
 +
incarnation: 12041451716771642396
 +
, name: "/device:GPU:0"
 +
device_type: "GPU"
 +
memory_limit: 2914163099
 +
locality {
 +
  bus_id: 1
 +
  links {
 +
  }
 +
}
 +
incarnation: 596105631149220834
 +
physical_device_desc: "device: 0, name: NVIDIA GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1"
 +
]
 
</source>
 
</source>
  
 
<br><br>
 
<br><br>

2022年4月14日 (四) 15:45的最后版本

[编辑] 1 Overview


TinyML 基金会在 2019 年组织了第一届峰会，这届峰会的成果如下：

  • TinyML 的技术硬件已经进入了实用性的阶段；
  • 算法，网络以及低于 100KB 的 ML 模型，已经取得重大突破；
  • 视觉，音频的低功耗需求快速增长。



[编辑] 2 Hardware




[编辑] 3 Quick Start 

$ pip3 install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow-gpu==2.5.0

$ python3
import tensorflow as tf
>>> print(tf.__version__)
2.5.0

>>> print('GPU: ', tf.test.is_gpu_available())
2022-04-01 03:24:13.099644: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1733] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: NVIDIA GeForce GTX 1050 computeCapability: 6.1
coreClock: 1.493GHz coreCount: 5 deviceMemorySize: 4.00GiB deviceMemoryBandwidth: 104.43GiB/s
2022-04-01 03:24:13.099980: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1871] Adding visible gpu devices: 0
2022-04-01 03:24:13.100248: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1258] Device interconnect StreamExecutor with strength 1 edge matrix:
2022-04-01 03:24:13.100459: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1264]      0
2022-04-01 03:24:13.100669: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1277] 0:   N
2022-04-01 03:24:13.100967: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1418] Created TensorFlow device (/device:GPU:0 with 2779 MB memory) -> physical G
PU (device: 0, name: NVIDIA GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1)
GPU:  True

>>> hello = tf.constant('Hello, Tensorflow!')
>>> print(hello)
tf.Tensor(b'Hello, Tensorflow!', shape=(), dtype=string)

>>> from tensorflow.python.client import device_lib
>>> print(device_lib.list_local_devices())
[name: "/device:CPU:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 12041451716771642396
, name: "/device:GPU:0"
device_type: "GPU"
memory_limit: 2914163099
locality {
  bus_id: 1
  links {
  }
}
incarnation: 596105631149220834
physical_device_desc: "device: 0, name: NVIDIA GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1"
]



来自“http://wiki.jackslab.org/index.php?title=TinyML&oldid=15498”
个人工具
名字空间

变换
操作
导航
工具箱