Exowatt fine-tunes Qwen3 models on SageMaker, reports beating Claude on two tasks
Overview
Exowatt used Amazon SageMaker AI serverless model customization to fine-tune two Qwen3 models, according to an AWS Machine Learning Blog post.
The fine-tuned Qwen3-1.7B, for edge alarm triage, scored 0.911 macro-F1, above Claude Sonnet 4.6's 0.698. The fine-tuned Qwen3-14B, for cloud energy-market analysis, scored 0.975, above Claude Opus 4.7's 0.625, at $0.60 per 1,000 answers versus $29.26 for Opus 4.7. The post says the pipeline is meant to produce task-specific models for edge and cloud deployment and that Exowatt can now run this path in days rather than months. These two models are the first outputs of that pipeline, not its end goal. The results are Exowatt's and AWS's reported figures; no independent evaluation is cited.
Written by AI from the articles below · updated Oct 27, 1:35 PM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- AWS Machine Learning BlogOfficialPickExowatt uses SageMaker serverless fine-tuning to beat frontier models on alarm triage and energy analysis
AIExowatt used Amazon SageMaker AI serverless model customization to fine-tune Qwen3-1.7B for edge alarm triage and Qwen3-14B for cloud energy-market analysis. The fine-tuned Qwen3-1.7B scored 0.911 macro-F1 on alarm triage, above Claude Sonnet 4.6's 0.698, and the fine-tuned Qwen3-14B scored 0.975 on energy-market analysis, above Claude Opus 4.7's 0.625. The Qwen3-14B cost $0.60 per 1,000 answers versus $29.26 for Claude Opus 4.7, and the post describes deploying to edge CPUs and Amazon Bedrock Custom Model Import.
Heat Trend
- Current comparable heat
- 6
- Peak comparable heat
- 10Oct 27, 2:00 PM ET
- Change in past 24 hours
- –
The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.