LLM EXPLORER 68,192 MODELS INDEXED

Qwen3.8 27B Coder390 FP8 DFlash2 V15 by LTM-Lab

By LTM-Lab · 6 downloads

Qwen3.8 27B Coder390 FP8 DFlash2 V15 is an open-source language model by LTM-Lab. Features: 27b LLM, License: apache-2.0, LLM Explorer Score: 0.26.

Base model:finetune:nerkyor/qw... Base model:nerkyor/qwen3.8-27b...   Conversational   Custom code   Dflash2   Fp8   Image-text-to-text   Mtp   Ninfer   Qwen3 5   Region:us   V15-episodic-memory   Vision

Qwen3.8 27B Coder390 FP8 DFlash2 V15 Parameters and Internals

LLM NameQwen3.8 27B Coder390 FP8 DFlash2 V15
Repository πŸ€—https://huggingface.co/LTM-Lab/Qwen3.8-27B-Coder390-FP8-DFlash2-v15 
Model NameMerkyor/Qwen3.8-27B-Coder390-EfficientThink
Base Model(s)  nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-MTP-DFlash2   nerkyor/Qwen3.8-27B-Coder390-EfficientThink-Opus5.5-GPT6Astra-Grok4.7-DSV4Pro-K3-SFT-RLOO-MTP-DFlash2
Model Size27b
Updated2026-11-21
MaintainerLTM-Lab
Model Typeqwen3_5
Model ArchitectureQwen3_5V15ForConditionalGeneration
Licenseapache-2.0
Model Max Length262144
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Torch Data Typebfloat16
Errorsreplace

Best Alternatives to Qwen3.8 27B Coder390 FP8 DFlash2 V15

Best Alternatives
Context / RAM
Downloads
Likes
...TRA HERETIC Uncensored V15 FP80K / 31 GB1240
...9 ULTRA HERETIC Uncensored V150K / 55.4 GB680
Note: green Score (e.g. "73.2") means that the model is better than LTM-Lab/Qwen3.8-27B-Coder390-FP8-DFlash2-v15.