Telescopium-Acyclic-Qwen3-0.6B-GGUF/README.md

license, language, base_model, pipeline_tag, library_name, tags

license	language	base_model	pipeline_tag	library_name	tags
apache-2.0	en	prithivMLmods/Telescopium-Acyclic-Qwen3-0.6B	text-generation	transformers	text-generation-inference

Telescopium-Acyclic-Qwen3-0.6B-GGUF

Telescopium-Acyclic-Qwen3-0.6B is a high-efficiency, multi-domain AI model fine-tuned on Qwen-0.6B that specializes in symbolic reasoning across mathematics, programming, and scientific domains. Leveraging a Directed Acyclic Graph (DAG) based reasoning methodology inspired by deepseek-r1 traces, it excels in decomposing complex problems into logical multi-step solutions. Featuring enhanced code expert clusters and an open code reasoning dataset, the model delivers unified outputs in LaTeX, Markdown, and various structured formats. Optimized for deployment on mid-range GPUs and edge AI systems, Telescopium-Acyclic-Qwen3-0.6B empowers developers, educators, and researchers with precise, step-by-step analytical reasoning for STEM tasks, algorithm synthesis, and technical documentation, while maintaining a lightweight footprint for versatile offline and cluster environments.

Model Files

File Name	Quant Type	File Size
Telescopium-Acyclic-Qwen3-0.6B.BF16.gguf	BF16	1.2 GB
Telescopium-Acyclic-Qwen3-0.6B.F16.gguf	F16	1.2 GB
Telescopium-Acyclic-Qwen3-0.6B.F32.gguf	F32	2.39 GB
Telescopium-Acyclic-Qwen3-0.6B.Q2_K.gguf	Q2_K	296 MB
Telescopium-Acyclic-Qwen3-0.6B.Q3_K_L.gguf	Q3_K_L	368 MB
Telescopium-Acyclic-Qwen3-0.6B.Q3_K_M.gguf	Q3_K_M	347 MB
Telescopium-Acyclic-Qwen3-0.6B.Q3_K_S.gguf	Q3_K_S	323 MB
Telescopium-Acyclic-Qwen3-0.6B.Q4_K_M.gguf	Q4_K_M	397 MB
Telescopium-Acyclic-Qwen3-0.6B.Q4_K_S.gguf	Q4_K_S	383 MB
Telescopium-Acyclic-Qwen3-0.6B.Q5_K_M.gguf	Q5_K_M	444 MB
Telescopium-Acyclic-Qwen3-0.6B.Q5_K_S.gguf	Q5_K_S	437 MB
Telescopium-Acyclic-Qwen3-0.6B.Q6_K.gguf	Q6_K	495 MB
Telescopium-Acyclic-Qwen3-0.6B.Q8_0.gguf	Q8_0	639 MB

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):