This is a remote position.

We are seeking a motivated and technically strong LLM Fine-Tuning and Data Engineering Intern to support our AI initiatives focused on large language model training and optimization.

The selected candidate will assist in preparing high-quality datasets, structuring instruction-tuning formats, and supporting fine-tuning workflows using modern open-source LLM frameworks.

This internship is ideal for candidates who want hands-on exposure to real-world LLM training pipelines, data curation processes, and model evaluation techniques.

Requirements

1. Data Preparation and Cleaning

---

• Clean, normalize, and pre process raw text datasets
• Remove duplicate, irrelevant, or low-quality entries
• Structure datasets into JSON/JSONL formats suitable for training
• Prepare instruction–response formatted datasets
• Maintain proper train/validation splits

2. Tokenization and Data Structuring

---

• Apply appropriate tokenizer for specific base models
• Analyze token length distributions
• Implement truncation and padding strategies
• Ensure dataset compatibility with sequence length constraints

3. Support Fine-Tuning Workflows

---

• Assist in setting up LoRA / QLoRA / PEFT-based fine-tuning
• Configure training parameters under supervision
• Monitor training metrics such as loss curves
• Maintain organized experiment records

4. Evaluation and Quality Assurance

---

• Conduct prompt-based evaluation of fine-tuned models
• Compare baseline vs fine-tuned model outputs
• Document findings and suggest improvements

5. Documentation and Reporting

---

• Maintain clear documentation of datasets and experiments
• Prepare structured reports on data quality and training outcomes
• Ensure reproducibility of experiments

Benefits

Internship Details

• Duration: 3–6 months
• Mode: Remote / Hybrid (as applicable)
• Stipend: As per industry standards and candidate capability
• Opportunity for full-time role based on performance​