Views
No views yet
Esper 3 is a specialist model built on Qwen 3, designed for coding, architecture, and DevOps reasoning. It has been fine-tuned using our proprietary DevOps, architecture, and code reasoning dataset generated with Deepseek R1. This tuning enhances its general and creative reasoning abilities, making it effective not only in problem-solving but also in general conversation. With its small model sizes, Esper 3 is optimized for fast inference, making it suitable for deployment on local desktops, mobile devices, and high-speed server environments.
| Filename | Size | Format | Description |
|---|---|---|---|
| Qwen3-4B-Esper3.BF16.gguf | 8.05 GB | BF16 | Brain Float 16-bit quantization |
| Qwen3-4B-Esper3.F16.gguf | 8.05 GB | F16 | Half precision (16-bit) floating point |
| Qwen3-4B-Esper3.F32.gguf | 16.1 GB | F32 | Full precision (32-bit) floating point |
| Qwen3-4B-Esper3.Q2_K.gguf | 1.67 GB | Q2_K | 2-bit quantization with K-quant |
| Qwen3-4B-Esper3.Q3_K_L.gguf | 2.24 GB | Q3_K_L | 3-bit quantization (Large) with K-quant |
| Qwen3-4B-Esper3.Q3_K_M.gguf | 2.08 GB | Q3_K_M | 3-bit quantization (Medium) with K-quant |
| Qwen3-4B-Esper3.Q3_K_S.gguf | 1.89 GB | Q3_K_S | 3-bit quantization (Small) with K-quant |
| Qwen3-4B-Esper3.Q4_K_M.gguf | 2.5 GB | Q4_K_M | 4-bit quantization (Medium) with K-quant |
| Qwen3-4B-Esper3.Q4_K_S.gguf | 2.38 GB | Q4_K_S | 4-bit quantization (Small) with K-quant |
| Qwen3-4B-Esper3.Q5_K_M.gguf | 2.89 GB | Q5_K_M | 5-bit quantization (Medium) with K-quant |
| Qwen3-4B-Esper3.Q5_K_S.gguf | 2.82 GB | Q5_K_S | 5-bit quantization (Small) with K-quant |
| Qwen3-4B-Esper3.Q6_K.gguf | 3.31 GB | Q6_K | 6-bit quantization with K-quant |
| Qwen3-4B-Esper3.Q8_0.gguf | 4.28 GB | Q8_0 | 8-bit quantization |
