udaylunawat/whats-this-rock

Telegram Rock Classification Bot!

Hi, I'm Uday Lunawat

Data Science | Software Developer | Lifelong learner

👨🏻‍💻  About Me

💡  I like to explore new technologies, develop software solutions and quick hacks.

💬  Feel free to reach out to me for pro bono consulting and volunteering, or just for some interesting discussion.
✉️  You can shoot me an email at [email protected]! I'll try to respond as soon as I can.
📄  Please have a look at my Website for more details about me. I'm open to feedback and suggestions!

Who Am I?

Data Wizard 🧙‍♂️

I am a Machine learning and Data Science practitioner who loves to uncover the hidden facts and meaningful insights from messy data. I'm a Machine Learning Engineer graduate from AAIC (Applied AI Course) where I learned building effective Machine Learning & Deep Learning Models, running Data Pipelines, Natural Language Processing, Image Processing, and deploying solutions to the cloud.

I graduated in Information Tech from RCOEM, Nagpur & since then I am working with people across the globe as a Machine Learning Engineer where I am learning new things in Data Science and helping people develop solutions to their data problems.

I was introduced to the field of Data Science and Machine learning during my 3rd year in College, during which I implemented two projects in Natural Language Processing. Since then, I breathe data.

In the past, I have worked with Tata Consultancy Services (TCS) and Infosys as a Business Intelligence and Python Intern.

What do I believe?

I am working towards a role where I can engage with a group of like-minded people who dream to help people all over the world make their lives easier with data.

🛠  Tech Stack 🔥

Languages:-

Data Analysis:-

Tools & Platforms:-

Hardware:-

Dev Environment:-

My Projects

🎁 Projects	🤖 Colab	⭐ Stars	📚 Forks	🛎 Issues	📬 Pull requests
What's this Rock - Telegram	Run & Deploy
Indian ALPR System - GCP	Run & Deploy
Covid-19 Classifier & Dashboard	Run & Deploy

My latest projects

• Whats-this-Rock
  An Image Classification pipeline for identifying rocks deployed as a Telegram Bot

• ALPR-India
  An end-end Deep Learning ALPR project on which you can play around with opencv morphological options and OCR

• Covid-19 Classifier & Dashboard
  An end-end Deep Learning based Covid-19 CT-scan Classification project with 97% accuracy and live Dashboard.

How to reach me 📱

Known on the internet as udaylunawat

😄 Know More About Me:

Visit to see my education, experience, projects, certifications, achievements and testimonials given by my clients.

🔖 Discover more about me at my Website

💼 Hire Me Resume

Watch my contribution graph eaten by the snake🐍

Show some ❤️ by starring some of the repositories!

👀 Visitors. Wow!

whats-this-rock's Issues

Linting using GItHub super linter

I am trying to understand the GIthub super linter and have added it to the Github actions workflow.
Need help understanding super linter and best linting practices.

Logging and warnings

Add proper logging and warnings, instead of using print statements.

Project Description

Am I permitted to make a PR for a little introduction to the project to serve as description in the main readME

Testing out resnet with different LR schedulers

Notion page with resources

• Learning rate schedulers

LR Schedulers to try

• Steps
• Cosine with restarts
• Classic cosine annealing (per epoch)
• Classic cosine annealing (per batch)
• Classic cosine annealing (per batch) with warmup

Inspiration Tweet

Sebastian Raschka's Tweet

Migrate use of ImageDataGenerator to keras_cv

Hey there @udaylunawat! Cool project you have going here!

I wanted to let you know that the Keras team has launched an extensive set of data augmentation layers in the new KerasCV repo: https://github.com/keras-team/keras-cv

I see you use ImageDataGenerator. May I recommend substituting this with RandAugment, CutMix, and MixUp? You can see how to do this here:

https://keras.io/guides/keras_cv/cut_mix_mix_up_and_rand_augment/

Thanks and let me know if you end up doing this!

ConvNextTiny not working

ConvNextTiny
Similar Issue on Tensorflow

Tensorflow Version:- 2.10

python src/models/train.py wandb.project=Whats-this-rockv7 \
                            wandb.mode=offline \
                            backbone=convnexttiny \
                            trainable=False \

error log

Found 2099 files belonging to 7 classes.
Found 447 files belonging to 7 classes.
Found 456 files belonging to 7 classes.
WARNING:tensorflow:Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
[2022-10-01 20:21:42,893][tensorflow][WARNING] - Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:42,918][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:42,946][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
[2022-10-01 20:21:42,967][tensorflow][WARNING] - Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
[2022-10-01 20:21:43,236][tensorflow][WARNING] - Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
[2022-10-01 20:21:43,335][tensorflow][WARNING] - Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:43,359][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:43,386][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
[2022-10-01 20:21:43,407][tensorflow][WARNING] - Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
[2022-10-01 20:21:43,582][tensorflow][WARNING] - Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
[2022-10-01 20:21:43,746][tensorflow][WARNING] - Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:43,778][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:43,812][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
[2022-10-01 20:21:43,845][tensorflow][WARNING] - Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
[2022-10-01 20:21:44,127][tensorflow][WARNING] - Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
[2022-10-01 20:21:44,230][tensorflow][WARNING] - Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:44,261][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:44,294][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
[2022-10-01 20:21:44,321][tensorflow][WARNING] - Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
[2022-10-01 20:21:44,507][tensorflow][WARNING] - Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
[2022-10-01 20:21:44,864][tensorflow][WARNING] - Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:44,888][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:44,914][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
[2022-10-01 20:21:44,935][tensorflow][WARNING] - Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
[2022-10-01 20:21:45,205][tensorflow][WARNING] - Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
[2022-10-01 20:21:45,303][tensorflow][WARNING] - Using a while_loop for converting RngReadAndSkip cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:45,328][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting Bitcast cause there is no registered converter for this op.
[2022-10-01 20:21:45,358][tensorflow][WARNING] - Using a while_loop for converting Bitcast cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
[2022-10-01 20:21:45,382][tensorflow][WARNING] - Using a while_loop for converting StatelessRandomUniformV2 cause there is no registered converter for this op.
WARNING:tensorflow:Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
[2022-10-01 20:21:45,556][tensorflow][WARNING] - Using a while_loop for converting ImageProjectiveTransformV3 cause there is no registered converter for this op.
Error executing job with overrides: ['wandb.project=Whats-this-rockv7', 'wandb.mode=offline', 'backbone=convnexttiny', 'trainable=False']
Traceback (most recent call last):
  File "/content/Whats-this-rock/src/models/models.py", line 43, in get_backbone
    input_shape=(cfg.image_size, cfg.image_size, cfg.image_channels),
  File "/usr/local/lib/python3.7/dist-packages/keras/applications/convnext.py", line 591, in ConvNeXtTiny
    classifier_activation=classifier_activation,
  File "/usr/local/lib/python3.7/dist-packages/keras/applications/convnext.py", line 521, in ConvNeXt
    )(x)
  File "/usr/local/lib/python3.7/dist-packages/keras/applications/convnext.py", line 287, in apply
    )(x)
  File "/usr/local/lib/python3.7/dist-packages/keras/utils/traceback_utils.py", line 70, in error_handler
    raise e.with_traceback(filtered_tb) from None
  File "/usr/local/lib/python3.7/dist-packages/tensorflow/python/framework/op_def_library.py", line 591, in _ExtractInputsAndAttrs
    f"{prefix} type "
TypeError: Exception encountered when calling layer "convnext_tiny_stage_0_block_0_layer_scale" (type LayerScale).

Input 'y' of 'Mul' Op has type float32 that does not match type float16 of argument 'x'.

Call arguments received by layer "convnext_tiny_stage_0_block_0_layer_scale" (type LayerScale):
  • x=tf.Tensor(shape=(None, 56, 56, 96), dtype=float16)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "src/models/train.py", line 219, in main
    model, history = train(cfg, train_dataset, val_dataset, class_weights)
  File "src/models/train.py", line 49, in train
    model = get_model(cfg)
  File "/content/Whats-this-rock/src/models/models.py", line 67, in get_model
    base_model = get_backbone(cfg)
  File "/content/Whats-this-rock/src/models/models.py", line 46, in get_backbone
    raise NotImplementedError("Not implemented for this backbone.")
NotImplementedError: Not implemented for this backbone.