Model Intelligence Sheet
ecwu/qwen35-4b-423333-lora-gguf-bf16 overview
Qwen3.5 4B 423333 LoRA merged GGUF This repository contains a GGUF export of ecwu/qwen35 4b unsloth 423333 , merged into unsloth/Qwen3.5 4B . Conversion notes:…
Runs locally from ~7.85 GB disk (8 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
1 GGUF files detected
Direct downloads for local inference
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| qwen35-4b-bf16.gguf | GGUF | BF16 | 7.85 GB | Download |
Model Details
Model README
---
language:
- zh
- en
base_model: unsloth/Qwen3.5-4B
tags:
- gguf
- qwen3.5
- lora
- merged
library_name: llama.cpp
---
Qwen3.5 4B 423333 LoRA merged GGUF
This repository contains a GGUF export of ecwu/qwen35-4b-unsloth-423333, merged into unsloth/Qwen3.5-4B.
Conversion notes:
- LoRA was merged with the Qwen3.5 conditional generation architecture, matching adapter keys under
model.language_model.*. - GGUF conversion used llama.cpp commit
4fc4ec554. - GGUF metadata was corrected to
qwen35.block_count = 32andqwen35.nextn_predict_layers = 0, because the exported language model does not include MTP tensors.
Source adapter: https://huggingface.co/ecwu/qwen35-4b-unsloth-423333
Base model: https://huggingface.co/unsloth/Qwen3.5-4B
File: qwen35-4b-bf16.gguf
Quantization: BF16
Run ecwu/qwen35-4b-423333-lora-gguf-bf16 with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models