---
title: "Fine-Tuning"
description: "9 posts about Fine-Tuning, summarised, each linking to the original."
---

# Fine-Tuning
> 9 posts about Fine-Tuning, summarised, each linking to the original.

## Articles

### [How FactSet cut model training and deployment from days to hours with Databricks AI Runtime](https://yomu.fyi/post/how-factset-cut-model-training-and-deployment-from-days-to-hours-with.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Nikhil Hiriyur Sunderraj, Dakshina Murthy Malladi, Nikitha Goram, Ryan Roser, Kate Stepp, Julia Powell, Tejas Sundaresan, Lucas de Paula
- Published: Oct 15, 2026
- Type: Problem & solution

FactSet’s Text-to-Formula (TTF) system converts natural-language investment queries from professionals into correct FactSet Query Language (FQL) formulas. Initially built on retrieval-augmented generation (RAG), the approach hit a quality ceiling as queries became more specialized and formulas more complex. FactSet adopted fine-tuned open models within a Compound AI architecture, reducing end-to-end latency by about 60 percent and improving quality for complex formulas. Fine-tuning runs on Databricks AI Runtime with serverless GPU infrastructure, where engineers choose compute based on prompt length and output needs, and the pipeline integrates with MLflow, the Lakehouse, and Lakeflow Jobs for a unified workflow. This setup cut model setup, training, and deployment from days to hours, enabling the team to retrain every two to three weeks as new data and models become available.


### [How Flo Health built safer, more useful open models with Databricks AI Runtime](https://yomu.fyi/post/how-flo-health-built-safer-more-useful-open-models-with-databricks-ai.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Vladislav Nedosekin, András Meczner, Julia Powell, Tejas Sundaresan, Michael Shtelma
- Published: Oct 12, 2026
- Type: Problem & solution

Flo Health, a women’s health app with 81 million monthly active users, aimed to improve generative AI responses for medical safety and usefulness while operating at scale. Relying on general-purpose models created cost, dependency, and domain-expertise limitations, so Flo built a post-training pipeline with Databricks AI Runtime. The team defined 26 AI judges across three categories: 12 medical-safety judges acting as hard constraints, 6 micro judges for nuanced behaviors, and 8 usefulness criteria. Using these judges as a reward signal, they generated over 15,000 synthetic questions from privacy-first principles and applied reinforcement fine-tuning with supervised fine-tuning (SFT) and direct preference optimization (DPO). After fine-tuning on 32 NVIDIA H100 GPUs using PyTorch and Fully Sharded Data Parallel (FSDP), medical safety scores rose from 34% to 61%, usefulness from 42% to 76%, and the combined metric improved nearly threefold to 47%.


### [Fine-tuning LLMs for risk assessment](https://yomu.fyi/post/fine-tuning-llms-for-risk-assessment.md)
- Company: [Stripe](https://yomu.fyi/company/stripe.md)
- Author: Hangzhi Guo
- Published: Oct 9, 2026
- Type: Problem & solution

Stripe makes thousands of near real-time risk decisions every second, balancing fraud detection against false positives that disrupt legitimate businesses. Zero-shot prompting achieved 0.487 precision, domain-guided prompting 0.706, and human reviewers 0.790 at the same recall. The team applied supervised fine-tuning (SFT) to an open-weight 4B-parameter LLM, serializing features like card payments and account activity into text records, training it on labels, and deriving a continuous fraud score from the next-token probabilities normalized over two label tokens. SFT increased average precision by approximately 24% over domain-guided prompting and 70% over zero-shot prompting, and consistently outperformed XGBoost across multiple risk tasks. Linear probing of hidden states showed that SFT made benign signals more linearly decodable, with ROC-AUC improving from 0.583 to 0.642, indicating the model better recognizes legitimate business activity and reduces false positives.


### [The model that didn't exist, so you made it yourself](https://yomu.fyi/post/the-model-that-didn-t-exist-so-you-made-it-yourself.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: yuvraj sharma, Abubakar Abid
- Published: Oct 8, 2026
- Type: Tutorial

The author wanted a small version of the prompt rewriter that ships with Qwen-Image 2.1, but the official 9B model requires about 20 GB of memory. Using ML Intern in HuggingChat, they created a 0.8B version that runs on a CPU, returns valid output 99.7% of the time, and uses about a quarter of the teacher’s tokens. The compute for the entire project, including having the 9B model label 8,797 example requests, came to USD 16. Over a few days, the author made five more models using the same approach, each starting as a message in HuggingChat with ML Intern switched on. The author’s prompts include verified facts, a baseline request, a smoke test, and a spending cap. The six models built include a citrus disease vision model, a Huggy character LoRA, a camera-angle LoRA, a doodle-in LoRA, a pocket rewriter, and a distilled text-to-image model.


### [One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO](https://yomu.fyi/post/one-model-family-two-gold-level-results-fine-tuning-nemotron-for-ioi-a.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: Aleksander, Igor Gitman, Sean Narenthiran, Mehrzad Samadi, Somshubra Majumdar
- Published: Oct 7, 2026
- Type: Announcement

Starting from Nemotron 3, the team applied supervised fine-tuning (SFT) and reinforcement learning (RL) to create specialists for competitive programming and olympiad mathematics. At the IOI, the Nemotron-3-Ultra-CC model, trained with SFT on 22,000 curated problems and paired with GenCorrect iterative refinement, scored 535.4 of 600 points, exceeding both the gold threshold and the top human score. At the IMO, a combination of SFT and RL checkpoints scored 30 of 42 points by generating, critiquing, and refining natural-language proofs without external tools. Both projects followed a shared recipe: start with a strong base, curate domain-specific data, apply standard post-training, and tie the model to an inference loop that improves candidates. The results show that specialization and test-time compute together produce gold-level performance, and all models, data, and inference pipelines are available on Hugging Face.


### [Open-sourcing AstaBrief, the fast report-generation model in Asta](https://yomu.fyi/post/open-sourcing-astabrief-the-fast-report-generation-model-in-asta.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: Kyle Wiggers
- Published: Oct 2, 2026
- Type: Announcement

AstaBrief is an open-weights 8B-parameter model developed by Ai2 for generating cited scientific reports from a research question and literature excerpts, designed for evidence-grounded outputs. It was trained from Qwen3-8B using supervised fine-tuning on 47K filtered real user queries and direct preference optimization, focusing on data quality. Unlike Asta’s Thinking mode, which uses a multi-step pipeline, AstaBrief writes the full report in a single pass to reduce generation time. In Asta’s Fast mode, it averages 51.1 seconds per report versus 178.5 seconds for Thinking mode, roughly 3.5× faster, while competitive on answer and citation quality. The model and training data are open-sourced on Hugging Face, enabling institutions to deploy behind their own firewall and supporting future open scientific models.


### [AutoSynthData: Generating Training Data for Enterprise Agents](https://yomu.fyi/post/autosynthdata-generating-training-data-for-enterprise-agents.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: Esakkivel Esakkiraja, Shruthan Radhakrishna, Denis Akhiyarov, Sagar Davasam
- Published: Oct 2, 2026
- Type: Problem & solution

AutoSynthData, built by ServiceNow CoreAI, generates training data for enterprise agents by targeting capability gaps from evaluation, using a stronger teacher such as Qwen3.8-27B to demonstrate successes. It uses a target model’s failures and a stronger teacher’s successes to decide what to learn next. The system produces tasks composing a system specification, user prompt, and verifier, ensuring feasibility, realism, and difficulty against the target environment. After sample-level checks and batch-level review, the dataset grows via target and multiply phases, generating 2,000 samples in 18 hours for the Hybrid domain. In EnterpriseOps Gym, synthetic SFT on Gemma-4-26B-A4B-it improved mean Pass@1 by 7.2 points (35% relative) in Hybrid and from 18.77% to 27.18% in ITSM, showing the approach works across domains.


### [On the Effectiveness-Fluency Trade-Off in LLM Conditioning: A Systematic Study](https://yomu.fyi/post/on-the-effectiveness-fluency-trade-off-in-llm-conditioning-a-systemati.md)
- Company: [Apple](https://yomu.fyi/company/apple.md)
- Published: Sep 30, 2026
- Type: Benchmark

Controlling the output of Large Language Models (LLMs) is a central challenge for their reliable deployment, yet current conditioning methods are often evaluated narrowly on effectiveness without considering generation quality. This study systematically investigates a range of conditioning methods in both concept injection and removal scenarios, including efficient steering, activation steering, prompting, and supervised fine-tuning. The authors find that efficient steering methods frequently achieve their conditioning goal at a steep cost to fluency, degrading generation quality. They identify a critical previously overlooked interaction: activation steering methods are far less effective on instruction-tuned models than on their base counterparts. Simple prompting and supervised fine-tuning are viable options for concept injection but are not as good at concept removal, while cheaply computed textual metrics highly correlate with costly LLM-as-judge scores, providing practical insights for evaluation.


### [Together AI expands fine-tuning service with more models, live metrics, and finer controls](https://yomu.fyi/post/together-ai-expands-fine-tuning-service-with-more-models-live-metrics.md)
- Company: [Together AI](https://yomu.fyi/company/together-ai.md)
- Published: Sep 11, 2026
- Type: Announcement

Together AI is expanding its Fine-Tuning service to cover newer open-weight models and more of the workflow from dataset inspection through model selection. The update adds live training and evaluation metrics through the API, CLI, dashboard, and Python SDK, including curves that can be compared across in-flight jobs and underlying series that can be exported. Training controls now include Expert LoRA for mixture-of-experts layers, early stopping based on validation loss, and arbitrary effective batch sizes through gradient accumulation. The post reports that Expert LoRA recalled up to 89% of 200 invented facts versus 15% for attention-only adapters, while listed LoRA training prices fall by 30% to 70% for selected models, including the gpt-oss series.