Fine-Tuning Engineer Mid-Senior · Full-time · Remote →

Thewebvale; Bhopal; Hybrid; INR 1,200,000 - 2,400,000; Full time

Benefits offered by this job

Remote-first

Bhopal hybrid studio

Client-facing projects

Founders exposure

Learning & tooling support

Portfolio showcase

Job summary

TheWebVale is seeking a Fine-Tuning Engineer to tailor open-source foundation models for client domains, through instruction tuning, dataset prep, and parameter optimization. You’ll adapt models like Llama, Mistral, and Qwen for legal parsing, clinical terminology, and education tutoring, working with AI research engineers and backend teams.

This remote-first role offers a hybrid setup with a Bhopal studio, real client products across industries, and direct collaboration with the founder and

Qualifications

• Possess 3–5 years of software or machine learning experience with dedicated fine-tuning work.
• Demonstrate deep proficiency with Python, PyTorch, Hugging Face Transformers, and PEFT.
• Show practical experience fine-tuning models on GPU cloud instances with CUDA.
• Understand tokenization nuances, padding strategies, loss functions, and weight merging.
• Possess disciplined habits tracking experiments using tools like Weights & Biases or MLflow.
• Communicate model performance gains and hardware trade-offs clearly to team leaders.

Responsibilities

• Prepare, format, and tokenize instruction datasets for supervised fine-tuning (SFT).
• Fine-tune open-weight models using PEFT techniques including LoRA, QLoRA, and full weights.
• Monitor training metrics, learning rate schedules, and loss curves to prevent overfitting.
• Evaluate fine-tuned models against baseline foundation weights across standard benchmarks.
• Quantize and export fine-tuned adapter weights for low-latency serving on local servers.
• Document training hyperparameters, compute budgets, and reproducibility steps.

Skills

Python

PyTorch

Hugging Face Transformers

PEFT

CUDA

Tools

Python

PyTorch

LoRA

Fine-Tuning

PEFT

CUDA

vLLM

Job description

You will adapt open-source language models to specialized client domains through instruction tuning, dataset preparation, and parameter optimization.

As a Fine-Tuning Engineer at TheWebVale, you will specialize open-weight foundation models to perform specific client business tasks. You will adapt models like Llama, Mistral, and Qwen for legal document parsing, clinic terminology triage, and customized educational tutoring.You will partner with AI research engineers and backend developers throughout our two-to-three week release cadences. You will balance model training loss with inference speed, deploying specialized weights that clients own and run privately.

Python

PyTorch

LoRA

Fine-Tuning

PEFT

CUDA

vLLM

03Responsibilities

Your day to day.

6 things, one at a time. Keep scrolling.

Prepare, format, and tokenize instruction datasets for supervised fine-tuning (SFT).

Fine-tune open-weight models using PEFT techniques including LoRA, QLoRA, and full weights.

Monitor training metrics, learning rate schedules, and loss curves to prevent overfitting.

Evaluate fine-tuned models against baseline foundation weights across standard benchmarks.

Quantize and export fine-tuned adapter weights for low-latency serving on local servers.

Document training hyperparameters, compute budgets, and reproducibility steps.

04Requirements

What we look for.

6 things, one at a time. Keep scrolling.

• 01 / 06 Possess 3–5 years of software or machine learning experience with dedicated fine-tuning work.
• 02 / 06 Demonstrate deep proficiency with Python, PyTorch, Hugging Face Transformers, and PEFT.
• 03 / 06 Show practical experience fine-tuning models on GPU cloud instances with CUDA.
• 04 / 06 Understand tokenization nuances, padding strategies, loss functions, and weight merging.
• 05 / 06 Possess disciplined habits tracking experiments using tools like Weights & Biases or MLflow.
• 06 / 06 Communicate model performance gains and hardware trade-offs clearly to team leaders.

What you get.

thewebvale.

Remote-first, with a Bhopal studio for hybrid days

Real client products across many industries

Direct work with the founder and senior team

Learning and tooling support

Your work featured in our public portfolio

06How we hire

Three steps, no tricks.

1. 1

   Intro chat

   A short call about what you want to work on and how you like to work.

2. 2

   Work review

   Walk us through projects, code or designs you are proud of.

3. 3

   Paid task

   A small paid task or pairing session, so both sides see the fit.