adaption-gpt-oss-120b-telugu-instruct

Adaption AutoScientist — Telugu General Capabilities (120B LoRA)

Powered by Adaptive Data — Adaption Labs

🎯 Core Problem Solved & Real-World Impact

Telugu is spoken by over 80 million people but remains severely underrepresented in modern LLM training data. Most large language models produce broken grammar, mixed-script outputs, or refuse to respond in Telugu entirely. This creates a digital divide where Telugu speakers cannot access AI-powered tools in their native language.

This model solves that by adapting a 120B-parameter model to achieve robust Telugu language competence across diverse tasks — from news summarization and QA to creative writing and translation — demonstrating that targeted, high-quality data adaptation via AutoScientist can transform a primarily English LLM into a capable Telugu-speaking assistant.

Model Details

AutoScientist Platform Integration

The model was adapted using the data-centric tools on the AutoScientist platform:

• Adaptive Data Pipeline: Cleaned Telugu text encoding, removed broken Unicode, and normalized script inconsistencies. This automatic data cleaning upgraded the quality score from Grade C (6.0) to Grade B (8.14), a +35.6% relative improvement.
• Prompt Deduplication & Rephrase: Deduplicated semantically equivalent Telugu questions and diversified prompt phrasing, improving conversational versatility.
• Reasoning Traces: Guided the model to output logical chain-of-thought steps in Telugu before presenting final summaries, translations, or answers.
• Hyperparameter Optimization: Orchestrated optimal LoRA hyperparameter search and managed the training loop for the 120B parameter base model.
• Evaluation: Evaluated our adapted model against the base model on unseen Telugu prompts, with the judge scoring based on the quality of language competence.

📊 Dataset & Quality Improvements

• Dataset Name: all_combined_general
• Dataset ID: 1abb2918-abc3-4581-907d-be1470787ffc
• Dataset Size: 77,653 rows
• HuggingFace: narendarcodes/Telugu-MultiTask-Instruct-77K

Composition & Source Licenses:

Overall License: CC-BY-SA-3.0 (respecting the most restrictive source)

Adaptive Quality Evaluation:

⚙️ Training Configuration

📈 Evaluation Results

LLM-as-a-Judge Win Rate: 66% adapted vs. 34% base model

The adapted model demonstrates significantly improved Telugu language competence, outperforming the base model on unseen Telugu prompts judged on grammar, vocabulary, and natural language quality.

Intended Use Cases

• Telugu news headline and article generation
• Question answering based on Telugu news and Andhra Pradesh facts
• Creative Telugu poetry completion
• English to Telugu translation
• Telugu instruction following and general assistance

How to Use

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base_model = AutoModelForCausalLM.from_pretrained("togethercomputer/gpt-oss-120b-bf16")
model = PeftModel.from_pretrained(base_model, "narendarcodes/adaption-gpt-oss-120b-telugu-instruct")
tokenizer = AutoTokenizer.from_pretrained("narendarcodes/adaption-gpt-oss-120b-telugu-instruct")

Citation

@misc{golla2026telugu,
  title={Adaption AutoScientist Telugu General Capabilities Model},
  author={Golla Narendar},
  year={2026},
  note={Trained using Adaption Labs AutoScientist platform. Powered by Adaptive Data.}
}

---

Powered by Adaptive Data — Adaption Labs

Built for the 2026 Adaption AutoScientist Challenge — Language Category

Framework Versions

• PEFT 0.15.1