Train Muse Glimmer For Free

Train Muse Glimmer for free: Our recipe for out-betting Jev

Join us for our inaugural conference, Forge 2026

Blog

Train Muse Glimmer For Free Train Muse Glimmer for free: Our recipe for out-betting Jev

PUBLISHED 10/20/2026

Table of Contents TL;DR Building the dataset One training pass, better predictions A little confidence goes a long way Would you bet on it? Faster decisions with System One What will you train Muse Glimmer to do?

Table of Contents

TL;DR

• Muse Glimmer 30B is free to train with supervised fine-tuning (SFT) through November 30 , kicking off our Free Weights Club. • Tuned Muse Glimmer outperformed Jev by 11.2 percentage points in prediction accuracy on unseen topics , reaching 58.3% versus 47.1%. In a historical backtest, it also turned a hypothetical $1,000 into $1,071 after modeled trading fees , while both Jev and base Glimmer lost money. • Fireworks’ optimized System One serving stack delivered 2-3× faster median server processing than standard BF16 chat completions in a separate synthetic benchmark. Both used the same adapter and a single B200 GPU; precision and runtime differed. • Try our experiment with the Kalshi cookbook . For your own projects, Managed Training runs the SFT job from your examples, while the Serverless Training API gives you control over the training loop.

Welcome to the Fireworks Free Weights Club . For a limited time, we’ll bring you new post-training research, a recipe you can build on, and free training on a featured model so you can run experiments of your own. First up: Muse Glimmer 30B, with SFT training free through November 30. It’s multimodal, offers strong performance for the cost, and makes a great starting point for personal agents. For our first experiment, we asked whether training Muse Glimmer on historical Kalshi markets could give it an edge in predicting outcomes. We evaluated its predictions against Jev and the original model on markets excluded from the training set. We call the version we trained tuned Glimmer and the original model base Glimmer . Tuned Glimmer came out ahead of both comparison models on prediction accuracy. In a historical backtest, it turned a hypothetical $1,000 into $1,071 after modeled fees, while both comparison models lost money. Here’s what we trained, how we tested it, and how you can try it yourself. Start training for free. Building the dataset

Jev gave us a model to beat; Kalshi gave us a way to keep score. Settled markets come with an answer key: take “Will the high in New York exceed 80°F?” Once the market settles, there’s an outcome to check the prediction against. The challenge is learning something that carries over to the next question. Each example contained the market’s title, description, rules, and date, with the settled outcome as its label. The model saw no market prices or news. We kept training, calibration, and evaluation separate: • Training: 20,051 questions from 6,461 events that settled before January 4, 2026. • Calibration: 6,232 questions from 2,000 events that opened January 11 or later and settled before April 1, 2026. • Evaluation: 33,712 questions from 8,765 events that opened April 8, 2026 or later.

Within the evaluation set, 3,567 questions came from 113 topic families excluded from both training and calibration . We report results for the full test set and this unseen-topic subset to examine how well Glimmer handled later events and unfamiliar topics. We removed near-duplicate inputs, rule edits made after markets opened, and later-added winning options that could reveal outcomes. Each test event was evaluated once. One training pass, better predictions

These questions already have a menu of possible answers. We used the Fireworks Training API with K-way cross-entropy loss to train Glimmer to assign higher probability to the correct outcome. The loss considers only the answer choices for that question, rather than the model’s full vocabulary. We trained a LoRA adapter using SFT, with one pass through the training data. On the full test set, tuned Glimmer reached 64.8% accuracy , compared with 56.4% for base Glimmer and 45.9% for Jev. On unseen topics, tuned Glimmer reached 58.3% , ahead of base Glimmer’s 53.2% and Jev’s 47.1% . That puts it 11.2 percentage points ahead of Jev on topic families it hadn’t encountered during training. The common-answer reference always chooses “No” for yes/no questions and the most frequent correct option position for multiple-choice questions. It uses the evaluation answers to establish that reference, so it provides context rather than representing a trained predictor. A little confidence goes a long way

A 51% prediction and a 90% prediction might both land on “Yes,” but we’d want to treat those calls differently. Before using tuned Glimmer’s probabilities to select and size bets, we calibrated its confidence against observed outcomes. We fitted a separate calibration temperature for each question type using the held-out calibration set. This rescales tuned Glimmer’s probabilities without retraining its weights. The goal: predictions assigned an 80% probability should come true about eight times out of ten. Only tuned Glimmer received this calibration step, base Glimmer received neither task-specific training nor calibration. We used Jev’s probabilities directly from its API without additional calibration, consistent with its intended out-of-the-box use. The chart shows accuracy minus confidence on the evaluation data. Closer to zero is better; negative values indicate overconfidence and positive values indicate underconfidence. Tuned Glimmer’s confidence wasn’t just bravado. When it assigned an answer at least 80% probability , it was right 86.8% of the time , both across the full test and within the unseen-topic subset. It was also selective: 14.4% of full-test questions and 6.1% of unseen-topic questions cleared that threshold. Filtering for confidence gave us fewer predictions, but a much stronger hit rate. There’s a pattern to those confident calls. Many were “No” answers to related threshold questions, such as whether a value would exceed 100, 110, or 120. Those outcomes often move together, so the correct predictions aren’t all independent successes. Would you bet on it?

A confident prediction still needs a worthwhile price. If a bet pays $1 when it wins and the model estimates an 80% chance of winning, its expected payout is 80¢. Our rule required the purchase price plus modeled fees to be at least 5¢...