🚀 Medial Secures Investment on Shark Tank India - Fueling the Future of Professional Social Networking. 🔥

News

Messages

Try our Valuation Calculator →

Back

Narendra

Willing to contribut... • 4m

I fine-tuned 3 models this week to understand why people fail. Used LLaMA-2-7B, Mistral-7B, and Phi-2. Different datasets. Different methods (full tuning vs LoRA vs QLoRA). Here's what I learned that nobody talks about: 1. Data quality > Data quantity (but no one checks quality) I ran the same fine-tuning job twice: First: 10,000 examples, auto-generated Second: 1,000 examples, manually curated The 1,000-example model performed 23% better on my test set. But here's the thing: I had to manually inspect samples to know this. Most tools just accept whatever CSV you upload. 2. LoRA isn't always the answer QLoRA used 75% less GPU memory than LoRA. But it was 66% slower and cost 40% more (because runtime). For my use case, LoRA was better. But the "conventional wisdom" says QLoRA for everything. The real answer: It depends on your data size, timeline, and budget. 3. Hyperparameter tuning isn't random—but it looks random I tested learning rates from 1e-5 to 1e-3. Below 5e-5: Training barely moved (10+ hours, no convergence) Above 5e-4: Loss exploded immediately Sweet spot: 5e-4 to 1e-4 But that's for this specific dataset and model combo. Every project needs its own tuning. Yet people copy parameters from tutorials and hope. What I'm building: A system that does this analysis automatically: → Scans your data → recommends LoRA vs QLoRA vs full tuning → Estimates training time + cost for each option → Runs mini-experiments (10% of data) to validate hyperparameters before full training Think of it as "pair programming" for fine-tuning. Piloting with 5 users next month. If you've struggled with this, let's talk. #MachineLearning #LLM #AI #FineTuning #BuildInPubl #DataScience

2 Replies

Replies (2)

Rahul Agarwal

Founder | Agentic AI... • 4m

Really insightful breakdown—especially the point about data quality over quantity. Excited to see how your system makes fine-tuning more efficient and accessible.

Amit Maurya

Transforming Ideas i... • 4m

are you there???

Recommendations from Medial

Aditya Karnam

Hey I am on Medial • 11m

"Just fine-tuned LLaMA 3.2 using Apple's MLX framework and it was a breeze! The speed and simplicity were unmatched. Here's the LoRA command I used to kick off training: ``` python lora.py \ --train \ --model 'mistralai/Mistral-7B-Instruct-v0.2' \ -

Varun reddy
•

GITAM • 1y

Fine-Tuning: The Secret Sauce of AI Magic! Ever wonder how AI gets so smart? It’s all about fine-tuning! Imagine a pre-trained model as a genius with general knowledge. 🧠✨ Fine-tuning takes that genius and hones its skills for a specific task, li

1 Reply

Kimiko

Startups | AI | info... • 9m

X updates its developer agreement to ban third parties from using the X API or X Content for training or fine-tuning foundation or frontier AI models.

AI Engineer

AI Deep Explorer | f... • 11m

"A Survey on Post-Training of Large Language Models" This paper systematically categorizes post-training into five major paradigms: 1. Fine-Tuning 2. Alignment 3. Reasoning Enhancement 4. Efficiency Optimization 5. Integration & Adaptation 1️⃣ Fin

Nikhil Raj Singh

Entrepreneur | Build... • 6m

Hiring AI/ML Engineer 🚀 Join us to shape the future of AI. Work hands-on with LLMs, transformers, and cutting-edge architectures. Drive breakthroughs in model training, fine-tuning, and deployment that directly influence product and research outcom

4 Replies

Anonymous

Hey I am on Medial • 8m

ppl just take open sourced LLM and do some fine-tuning to it and feed some of their own data and call it their own custom LLM. is it legit to do that?

2 Replies

Gigaversity

Gigaversity.in • 7m

Overfitting, underfitting, and fitting — these aren't just technical terms, but critical checkpoints in every machine learning workflow. Understanding these concepts is key to evaluating model behavior, improving generalization, and building solutio

AI Engineer

AI Deep Explorer | f... • 10m

LLM Post-Training: A Deep Dive into Reasoning LLMs This survey paper provides an in-depth examination of post-training methodologies in Large Language Models (LLMs) focusing on improving reasoning capabilities. While LLMs achieve strong performance

Vansh Khandelwal

Full Stack Web Devel... • 2m

Agentic AI—systems that act autonomously, are context-aware and learn in real time—paired with small language models (SLMs) enables efficient, specialized automation. SLMs are lightweight, fast, fine-tunable for niche domains and accessible to smalle

Yogesh Jamdade

..... • 1y

Hey everyone, I'm an engineering student geeking out over Generative AI. Loving LangChain, Hugging Face models, Crew.ai's chatbots, fine-tuning, and RAG. Plus, machine learning and data science are pretty cool too! Anyone else into this stuff? Looki