QVikhr-3-8B-Instruction-f32-GGUF

QVikhr-3-8B-Instruction is a powerful bilingual instruction-tuned language model based on the Qwen3-8B architecture, further refined with supervised fine-tuning (SFT) on the GrandMaster2 dataset, which is notable for its focus on Russian and English language data. This targeted training enables the model to generate precise, context-sensitive responses for a wide array of instruction-based tasks, such as answering questions, following detailed directions, and analyzing texts—particularly excelling at complex topics like mathematics and physics in Russian. By leveraging 8 billion parameters and advanced instructional alignment, QVikhr-3-8B-Instruction delivers near state-of-the-art results, matching or approaching the performance of larger models like GPT-4.1 on competitive benchmarks, while maintaining efficiency and versatility suitable for integration in both professional and user-facing applications.

Execute using Ollama

run ->

ollama run hf.co/prithivMLmods/QVikhr-3-8B-Instruction-f32-GGUF:BF16

Model Files

File Name	Quant Type	File Size
QVikhr-3-8B-Instruction.BF16.gguf	BF16	16.4 GB
QVikhr-3-8B-Instruction.F16.gguf	F16	16.4 GB
QVikhr-3-8B-Instruction.Q8_0.gguf	Q8_0	8.71 GB

Quants Usage

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):