Pawan.

LLM Fine tuning & Inference

[Pawan. @d0r1h]

Open Sourcing the Hindi New Summaries Dataset (100K+).

If you’re working on summarization models, multilingual NLP, or real-world language applications, this dataset is free to explore.

You can access the dataset here :

https://t.co/uiZ9GEaI9r

[Pawan. @d0r1h]

@RoboticswithV INDIA ROBOTS

[Pawan. @d0r1h]

@Ayussheth Yes, I want to get into the hardware.

[Pawan. @d0r1h]

@Narayani07 But I got a call for the interview at Cognizant.

[Pawan. @d0r1h]

I believe the only intern in history who would still be happy even if he doesn’t get converted into a PPO… is that Faridabad guy. 😄

[Pawan. @d0r1h]

Hey @real_deep_ml,

Is there a way to contribute problems on the platform.

I was working on OCR and Character Error Rate is way to evaluate OCR models.

I though to check on deep-ml platform but this problem is not there.

Let me know!

[Pawan. @d0r1h]

@ishq_me_ruswa Let me know if I'm fit for the role, attaching the resume for the same, happy to discuss further.

https://t.co/CmJMUr1ii1

[Pawan. @d0r1h]

Completed 13 days doing @real_deep_ml challenges daily.

Mostly I try to do in just Python, and NumPy in extreme cases.

Solution : https://t.co/nUG4AXOq4R

[Pawan. @d0r1h]

Day4 :: DeepML ::

Min-Max Normalization of Feature Values

Code : https://t.co/GWKZBnit6J

[Pawan. @d0r1h]

Day4 :: DeepML ::

Min-Max Normalization of Feature Values

Code : https://t.co/GWKZBnit6J

[Pawan. @d0r1h]

Came across this Bible on Fine tuning LLMs by people at @CeADARIreland and it covers everything from data prep to model training to Eval.

Excited to get in !

Link : https://t.co/CD0LBw6eju

[Pawan. @d0r1h]

Same with @DuckDuckGo , I think duck uses GPT-4o ...

[X Freeze @XFreeze]

2026 is not working great for Google 😑

[Pawan. @d0r1h]

Day4 :: DeepML ::

Min-Max Normalization of Feature Values

Code : https://t.co/GWKZBnit6J

[Pawan. @d0r1h]

I've pushed my code on GitHub:

https://t.co/4KLv5eJGh4

[Pawan. @d0r1h]

If you want to learn PyTorch & Deep Learning in 2026, I recommend this course by @mrdbourke.

I've going though it since new year eve and completed till Transfer Learning.

It's cover a lot in terms on Hands on in PyTorch.

Link : https://t.co/qv2hxSDYW5

[Pawan. @d0r1h]

Day3 :: Deep_ML ::

Linear Regression Using Normal Equation (OLS)

Code: https://t.co/0b36sXmHFf

[Pawan. @d0r1h]

Day2 :: Deep_ML :: Transpose of a Matrix

Code :
https://t.co/0b36sXnfuN

[Pawan. @d0r1h]

Day2 :: Deep_ML :: Transpose of a Matrix

Code :
https://t.co/0b36sXnfuN

[Pawan. @d0r1h]

@real_deep_ml I was solving this problem

https://t.co/LgujE3W1cD

Ran two test cases but failed the last because it was expecting the output to be in 3 decimal, I worked around using round(), but I think it should be mentioned in expected output (just a humble suggestions :)

[Pawan. @d0r1h]

@real_deep_ml No worries. Great Platform!

[Pawan. @d0r1h]

@real_deep_ml I was solving this problem

https://t.co/LgujE3W1cD

Ran two test cases but failed the last because it was expecting the output to be in 3 decimal, I worked around using round(), but I think it should be mentioned in expected output (just a humble suggestions :)

[Pawan. @d0r1h]

Have a Happy & successful 2026!

[Pawan. @d0r1h]

New Qwen models (Image Layered & Edit) by Alibaba is mind blowing, it gonna save ton of time of editors.

Thinking of building a simple app for the same which let's you edit the image using this.

Paper : https://t.co/oc5XOAV1jy