Downloads · 30 days
150
16% of all-time downloads
prithivMLmods/QVikhr-3-8B-Instruction-f32-GGUF
QVikhr-3-8B-Instruction-f32-GGUF is a text generation model from prithivMLmods. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as apache-2.0.
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…
Downloads · 30 days
150
16% of all-time downloads
All-time downloads
925
Public
Repo size
41.5 GB
Likes
2
Public
Click a slice to open those files.
.gguf41.5 GB · 100%
From the Hugging Face model README
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.
run ->
ollama run hf.co/prithivMLmods/QVikhr-3-8B-Instruction-f32-GGUF:BF16
| 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 |
(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):
