TensorFlow high-level APIs: Part 1

TensorFlow high-level APIs: Part 1 - loading data

Рет қаралды 121,810

Күн бұрын

Welcome to Part 1 of our mini-series on TensorFlow high-level APIs! In this 3 part mini-series, TensorFlow Engineering Manager Karmel Allison runs us through different scenarios using TensorFlow’s high-level APIs. Building a ML model takes a lot of time, effort, and often involves multiple stages. Luckily, TensorFlow high-level APIs aim to help you along with each stage, from the start of your idea, to training and serving large scale applications. Watch to discover the key steps in developing machine learning models, where TensorFlow comes in for each step, and lastly how to prepare and load your data!
Learn more about TensorFlow high-level APIs → bit.ly/2zETMOK
Want to watch more? → bit.ly/Coding-T...
Subscribe to the channel to catch new episodes of Coding TensorFlow → goo.gl/ht3WGe
And...stay tuned for Part 2 & 3!

Пікірлер: 75

@wrightw1982 4 жыл бұрын

Super helpful series. Doesn't just tell you how to build SOME PARTICULAR model, but how to build any basic/canned model. Best mini-tutorial I've seen so far.

@MikeDownes 5 жыл бұрын

Thanks this was incredibly well put together easy to understand and so so useful.

@bhesht 3 жыл бұрын

Immaculately described and concise at once. Thank you!

@mounkailagarba9952 7 ай бұрын

Very good presentation

@mdrayedbinwahed2172 5 жыл бұрын

Typo in 5:24. 'row_vals' should be 'vals'

@TerenceChill286 5 жыл бұрын

Thanks for that. I was wondering were the row_vals came from

@thinkforsociety 5 жыл бұрын

You are rite

@andrassteger9589 2 жыл бұрын

A lot of changes happened in the last 3 years. Maybe some update would be appreciated :-) Thank You!

@mounkailagarba9952 7 ай бұрын

Thank you very much

@soumyodey1795 5 жыл бұрын

There is a big mistake in the code shown in the video. The function _parse_csv_rows() is incorrect, it should be def _parse_csv_rows(*vals): soil_type_t = tf.convert_to_tensor(vals[14:54]) feat_vals = vals[:10] + (soil_type_t, vals[54]) features = dict(zip(col_names_features, feat_vals)) class_label = tf.argmax(vals[10:14], axis=0) labels = {col_names_label: class_label} return features, labels where col_names_features, col_names_label are col_names_features = [ "Elevation", "Aspect", "Slope", "Horizontal_Distance_To_Hydrology", "Vertical_Distance_To_Hydrology", "Horizontal_Distance_To_Roadways", "Hillshade_9am", "Hillshade_Noon", "Hillshade_3pm", "Horizontal_Distance_To_Fire_Points", "Soil_Type (40 binary columns)", "Cover_Type (7 types)" ] col_names_label = "Wilderness_Area (4 binary columns)"

@fahemhamou6170 2 жыл бұрын

تحياتي الخالصة شكرا جزيلا

@nunodias8927 5 жыл бұрын

Many things to do only for read a datafile. It will be more easy do a function to aggreate all possibilities and the end user set the options in according with your own needs. In fact this demo is useful only if you knows the structure of your data. If you can not open the file with a simple editor you don't know the structure of your data, so you can not setup TF to read the data if you do not know the structure of the data.

@gowthamkumar3296 4 жыл бұрын

In Tensorflow updated versions, numpy is removed from tensor, then how we will extract value from that tensor

@qiyang1247 4 жыл бұрын

What can we do to solve the problem of memory leaking when cooperating tf.keras with tf?

@nathanas64 4 жыл бұрын

Excellent presentation!

@tingnews7273 5 жыл бұрын

great vedio. Finally tensorflow make some vedio I can unstand. Good job

@TheIsrraaa 3 жыл бұрын

good vedio

@sachavanweeren9578 5 жыл бұрын

There is a bug in the code snippet on Parsing data - 'row_vals' is not defined, this should be replace by ' vals' . Better to share working code and provide links with the video.

@BadriNathJK 5 жыл бұрын

Can you tell me how to get each token vector when using both feature column and TF HUB? Because feature column gives average of token vector in a sentence. I am struck with the problem now

@neerajgobari9748 5 жыл бұрын

How can I use model final output to compute a prediction..so that I can compare prediction to ground truth value to compute loss function and fine tune my model

@write2sekhar 4 жыл бұрын

Good and well presented looking for coode and examples.

@simplysmile8573 4 жыл бұрын

github.com/adderbyte/Tensorflow_Tutorials--tf.Slim_GPU_included/blob/master/NewFeaturesTF/TF_DATA_API/Illustrtation_For_Tensorflow2.ipynb

@IgorSaprykin 5 жыл бұрын

Good one, Karmel!

@huqiao 5 жыл бұрын

This is not a very good example as you were using tf.contrib which will soon be deprecated in TensorFlow 2.0.

@shionwill7095 5 жыл бұрын

Please share your code and the data so that we can follow along this great tutorial!

@simplysmile8573 4 жыл бұрын

data example and code : github.com/adderbyte/Tensorflow_Tutorials--tf.Slim_GPU_included/blob/master/NewFeaturesTF/TF_DATA_API/Illustrtation_For_Tensorflow2.ipynb

@claudee8736 2 жыл бұрын

Building a ML Model is a multi-stage process = kzfaq.info/get/bejne/pax2la6gzL_Ilpc.html 1. Collect data, clean data and process data 2. Prototype and iterate on model arquitecture 3. Train and evaluate results 4. Prepare your model for production serving 5. Repeat all over again

@xkcd_68868 5 жыл бұрын

Is this for Tensorflow 2.0?

@AbhishekKumar-mq1tt 5 жыл бұрын

Thank u for this awesome video

@anubhav114 5 жыл бұрын

hello, thanks for the wonderful explanation of tensorflow usage. Could you please point me to the guide for i/o optimization @5:51 ?

@fusedglass01 4 жыл бұрын

Is this supposed to be TF 2.0? when I issue a print version I get "tensorflow_core._api.v2.version'". Unbelievable how much pain these guys cause, by changing stuff every release. Google... can you please Google the term "backwards compatibility?" I'm just trying to read in a two column csv. I agree it looked great while I was watching it, but its next to useless when I get errors about contrib and enable_eager arnt part of the tf module. Guys please release more of this style, but add more detail around what versions you need and at the end show us 100% of the code.

@aonoymousandy7467 5 жыл бұрын

my csv files have a string at the end of each row, and this is causing error in my code, how would I setup defaults to account that I have a row of 10 floats and the very last feature is a string?

@hixenP 2 ай бұрын

Karmel are you sitting down while making this video? JW 😊

@shaktirajsinhjadeja7386 4 жыл бұрын

Hi , Great video to learn from :) Will it be possible to get the access of dataset ?

@simplysmile8573 4 жыл бұрын

data example and code: github.com/adderbyte/Tensorflow_Tutorials--tf.Slim_GPU_included/blob/master/NewFeaturesTF/TF_DATA_API/Illustrtation_For_Tensorflow2.ipynb

@username42 5 жыл бұрын

where are the codes sources?

@deeplearningpartnership 5 жыл бұрын

Very interesting.

@eilaorielresearch1205 5 жыл бұрын

tf.contrib.data.CsvDataset is planned to be deprecated. the new API will be tf.data.experimental.CsvDataset

@JeremAl 5 жыл бұрын

Hi, great tutorial! Any chance to have the code? I am STRUGGLING to start with the data set!!! (8hours later... :( )

@willtun5385 5 жыл бұрын

SAME! ME WANT GITHUB CODE!

@simplysmile8573 4 жыл бұрын

data and code : github.com/adderbyte/Tensorflow_Tutorials--tf.Slim_GPU_included/blob/master/NewFeaturesTF/TF_DATA_API/Illustrtation_For_Tensorflow2.ipynb

@Mrduirk 2 жыл бұрын

un csv puede contener imagenes?

@kevinlieb 5 жыл бұрын

Please point us to the code and the demo data.

@RahulSingh-ln7bg 5 жыл бұрын

highly confusing demo

@kaan-tube 4 жыл бұрын

this probably assumes some level of python knowledge. if you don't know what a dictionary is or what the zip function do, I can see how this might get complicated.

@andreamarkos 4 жыл бұрын

how about a little bit of API sugar here?

@sushilchauhan7660 4 жыл бұрын

Good tutorial but no git hub link here to view code

@RARa12812 4 жыл бұрын

this is tensorflow 1.0 correct?

@mladenradosevic5333 5 жыл бұрын

mini-series on TensorFlow high-level APIs! In this 3 part mini-series is in wrong order ... first (I presume should be part 1, not part 3

@balajichetty305 5 жыл бұрын

Github code ?

@joryhansen50 4 жыл бұрын

why isn't contrib in TF2.0????

@obsiyoutube4828 4 жыл бұрын

please i need code and dataset?good presentation

@simplysmile8573 4 жыл бұрын

data and code example here : github.com/adderbyte/Tensorflow_Tutorials--tf.Slim_GPU_included/blob/master/NewFeaturesTF/TF_DATA_API/Illustrtation_For_Tensorflow2.ipynb

@meeexy116 3 жыл бұрын

tf.contrib is deprecated now.

@mdrayedbinwahed2172 5 жыл бұрын

Github?

@pabloe1802 5 жыл бұрын

Use tf.data.experimental.CsvDataset

@yonimalka6954 5 жыл бұрын

I have created a notebook to follow the video along if anyone is interested: gist.github.com/malkayo/71e483c6f234dd5aca0eaaa52a5c24e0

@chrischoir3594 5 жыл бұрын

Way too much work for something so basic

@stumpy1495 4 жыл бұрын

This should be removed or updated, there is no contrib within TF 2.x

@arkarnyanhein 5 жыл бұрын

you don't show step by step how you import the data gg. GG engineer

@ayasalama7965 5 жыл бұрын

Very confusing

@nandagopalgopakumar5626 5 жыл бұрын

FIRST :P

@zdenes 5 жыл бұрын

Please share your code and the data so that we can follow along this great tutorial!

@AZTECMAN 5 жыл бұрын

data: www.kaggle.com/uciml/forest-cover-type-dataset

@simplysmile8573 4 жыл бұрын

data and code : github.com/adderbyte/Tensorflow_Tutorials--tf.Slim_GPU_included/blob/master/NewFeaturesTF/TF_DATA_API/Illustrtation_For_Tensorflow2.ipynb

@johndoucette3687 4 жыл бұрын

@@simplysmile8573 This link does not seem to have anything to do with this video.

@simplysmile8573 4 жыл бұрын

@@johndoucette3687 it is the code implementation of what is discussed in the video. This might help follow the tutorial. I hope this helps.

@simplysmile8573 4 жыл бұрын

@@johndoucette3687 I see your point though. Well, the dataset that is being used might matter less than the actual computation and ability to reproduce the learning pipeline. It is not about forest coverage but about how to preprocess, analyse and build an algorithmic pipeline. The aim of the link is to show the pipeline that does similar thing (though the dataset is different). Feel free to use the forest coverage dataset to replace the dataset used in the link and it will be fine. Best Regards.