blossom-v4-qwen1.5-14B: benchmark results
A conversational instruction fine-tune released by Azure99 based on Qwen1.5-14B, using Azure99/blossom-chat-v2, Azure99/blossom-math-v3, Azure99/blossom-wizard-v2, Azure99/blossom-orca-v2. Training uses two stages; dataset example counts and epochs are not token or compute disclosures. Provider: Azure99. Access: Open.
Aggregate Rating 1417 ± 38, rank #5316 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - MMLU | 67.92 | Accuracy (%) (5-shot) | 90.2 |
| Open LLM Leaderboard v1 - GSM8K | 66.49 | Accuracy (%) (5-shot) | 81.8 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 55.21 | MC2 (%) (0-shot) | 64.5 |
| Open LLM Leaderboard v1 - HellaSwag | 79.84 | Normalized accuracy (%) (10-shot) | 39.3 |
| Open LLM Leaderboard v1 - ARC Challenge | 57.34 | Normalized accuracy (%) (25-shot) | 37.3 |
| Open LLM Leaderboard v1 - WinoGrande | 73.64 | Accuracy (%) (5-shot) | 32.9 |
Interactive version: theaggregate.ai/model?slug=blossom-v4-qwen1-5-14b · How It Works · Data refreshed daily, snapshot 2026-11-08.