---
title: "Swarit Shukla (@swaritshukla) | Hashnode"
description: "making machines think. Read the latest articles by Swarit Shukla on Hashnode."
url: "https://hashnode.com/@swaritshukla"
---

@swaritshukla

# Swarit Shukla

@swaritshukla · Joined April 2026

making machines think

### [GRPO: How Language Models Learn to Reason](https://swaritshukla.hashnode.dev/grpo-how-language-models-learn-to-reason?utm_source=hashnode&utm_medium=feed)

Do you remember the times when we used to make LLMs count the occurrence of a specific letter in a word, like "How many r's in strawberry?" Back then, LLMs used to get it wrong a lot of times, but now

5 min read · [swaritshukla.hashnode.dev](https://swaritshukla.hashnode.dev/grpo-how-language-models-learn-to-reason?utm_source=hashnode&utm_medium=feed)

[Sep 22](https://hashnode.com/posts/grpo-how-language-models-learn-to-reason/6ab22003c23b2ad2d81f0661)

### [Batch Size and LLM Inference Efficiency](https://swaritshukla.hashnode.dev/batch-size-and-llm-inference-efficiency?utm_source=hashnode&utm_medium=feed)

Having an optimal Batch size can decrease your models cost per token at the time of inference. This will be an explanation on how Batch size affects

4 min read · [swaritshukla.hashnode.dev](https://swaritshukla.hashnode.dev/batch-size-and-llm-inference-efficiency?utm_source=hashnode&utm_medium=feed)

[1](https://hashnode.com/posts/batch-size-and-llm-inference-efficiency/6a60ed36b9324f0f504e8b9c) [Jul 22](https://hashnode.com/posts/batch-size-and-llm-inference-efficiency/6a60ed36b9324f0f504e8b9c)

### [Nvidia Vs Everyone: Why everyone is building their own accelerators](https://swaritshukla.hashnode.dev/nvidia-vs-everyone-why-everyone-is-building-their-own-accelerators?utm_source=hashnode&utm_medium=feed)

There has been a race in the AI industry where companies are now preferring their own custom accelerators to train AI models and generate inference from them. But why? Why all of a sudden is everyone

7 min read · [swaritshukla.hashnode.dev](https://swaritshukla.hashnode.dev/nvidia-vs-everyone-why-everyone-is-building-their-own-accelerators?utm_source=hashnode&utm_medium=feed)

[May 4](https://hashnode.com/posts/nvidia-vs-everyone-why-everyone-is-building-their-own-accelerators/69f8409276c1469ba44247cf)

### [The Elegance of MoE: How Gemma 4’s 26B Model Runs Like a 4B Model](https://swaritshukla.hashnode.dev/the-elegance-of-moe-how-gemma-4-s-26b-model-runs-like-a-4b-model?utm_source=hashnode&utm_medium=feed)

How can a 26B model run like a 4B one? Discover the power of Mixture of Experts (MoE) in Google's latest open-source model release

5 min read · [swaritshukla.hashnode.dev](https://swaritshukla.hashnode.dev/the-elegance-of-moe-how-gemma-4-s-26b-model-runs-like-a-4b-model?utm_source=hashnode&utm_medium=feed)

[Apr 12](https://hashnode.com/posts/the-elegance-of-moe-how-gemma-4-s-26b-model-runs-like-a-4b-model/69db3cd0aadf1107e29b38f5)