Xerv-AI/MAXWELL
Text Generation
Params: 2B • Released: 10 days ago
MAXWELL is a fine-tuned, specialized variant of the Qwen2.5-Math-1.5B-Instruct architecture. It is optimized for high-precision analytical reasoning, mathematical computation, and physics problem-solving.
Internal Benchmarks
- MMLU-Pro
- 45.000
- GSM8K
- 70.000
- MATH-Hard
- 60.000
Xerv-AI/Ada
Text Generation
Params: 2B • Released: 13 days ago
Ada is an ultra-lightweight, high-speed, and highly optimized reasoning Small Language Model (SLM) derived from the powerful Qwen2.5-Math-1.5B architecture. Engineered specifically to bridge the gap between hyper-specialized graduate-level mathematical proofs and standard conversational utility.
Internal Benchmarks
- GSM8K
- 40.00%
- MATH
- 60.00%
- GRAD
- 40.00%
RainDrop-Demo-1
Zero-Shot Classification
Params: 68M • Released: Aug 13, 2025
Experimental zero-shot text classification model using a transformer-based dual-encoder architecture with contrastive learning. Early research/demo version trained on a small test-scale dataset. Not intended for production use.
Internal Benchmarks
- Note
- Modest accuracy reported qualitatively; no numerical results published