Llama 3.2 3B Instruct Abliterated needs about 3.7 GB of memory at Q4_K_M (a 2.2 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 13 quants, from 1.5 GB (Q2_K) to 7.2 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 3.8 GB.
Llama 3.2 3B Instruct Abliterated – a compact, instruction‑oriented LLM
Llama 3.2 3B Instruct Abliterated is a 3.61‑billion‑parameter model trained on 131072 context tokens. It is available in multiple GGUF quantizations, ranging from Q2_K to F16. The model is uncensored and designed for instruction following.
Good at
- Instruction following
- Low‑latency inference
- Compact quantizations
Limits
- No vision support
- Requires 8k context for estimates
- License is llama3.2
Best for
- Chatbots on edge devices
- Rapid prototyping of instruction models
- Educational demos
Who it’s for. Developers and researchers needing a lightweight Llama 3.2 variant for instruction tasks. The model is suitable for local deployment with limited GPU memory.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 1.5 GB | 2.9 GB | Fits | Get file → | |
| Q3_K_S | 1.7 GB | 3.2 GB | Fits | Get file → | |
| Q3_K_M | 1.9 GB | 3.3 GB | Fits | Get file → | |
| Q3_K_L | 2.0 GB | 3.4 GB | Fits | Get file → | |
| IQ4_XS | 2.1 GB | 3.5 GB | Fits | Get file → | |
| Q4_0 | 2.1 GB | 3.6 GB | Fits | Get file → | |
| Q4_K_S | 2.1 GB | 3.6 GB | Fits | Get file → | |
| Q4_K_M | 2.2 GB | 3.7 GB | Fits | Get file → | |
| Q5_K_S | 2.5 GB | 4.0 GB | Fits | Get file → | |
| Q5_K_M | 2.6 GB | 4.0 GB | Fits | Get file → | |
| Q6_K | 3.0 GB | 4.4 GB | Fits | Get file → | |
| Q8_0 | 3.8 GB | 5.3 GB | Fits | Get file → | |
| F16 | 7.2 GB | 8.7 GB | Too big | Get file → |
For 8 GB VRAM, Q4_K_M (2.24 GB) or Q5_K_S (2.54 GB) are good choices. If 4 GB VRAM is available, Q3_K_M (1.86 GB) or Q3_K_S (1.71 GB) work well. For maximum quality, use Q8_0 (3.84 GB) or F16 (7.22 GB) if memory allows.
What GPU do you need for Llama 3.2 3B Instruct Abliterated?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is Q5_K_M (2.6 GB file, about 4.0 GB in use).
- 6 GB VRAM: the best quant that fits is Q8_0 (3.8 GB file, about 5.3 GB in use).
- 8 GB VRAM: the best quant that fits is Q8_0 (3.8 GB file, about 5.3 GB in use).
- 12 GB VRAM: the best quant that fits is F16 (7.2 GB file, about 8.7 GB in use).
Which Llama 3.2 3B Instruct Abliterated quant should you choose?
Q4_K_M is the usual starting point for Llama 3.2 3B Instruct Abliterated: it balances file size and quality well. Here it is a 2.2 GB file.
The Q8_0 file is 3.8 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q2_K at 1.5 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Frequently asked questions about Llama 3.2 3B Instruct Abliterated
How much VRAM does Llama 3.2 3B Instruct Abliterated need?
About 3.7 GB at Q4_K_M with an 8K context window. That is the 2.2 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Llama 3.2 3B Instruct Abliterated on an 8 GB GPU?
Yes, with Q8_0: the 3.8 GB file needs about 5.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Llama 3.2 3B Instruct Abliterated GGUF?
It depends on the quant: Q2_K is 1.5 GB, Q3_K_S is 1.7 GB, Q3_K_M is 1.9 GB, Q3_K_L is 2.0 GB, IQ4_XS is 2.1 GB, Q4_0 is 2.1 GB and 7 more.
What context length does Llama 3.2 3B Instruct Abliterated support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.9 GB at 8K and 3.8 GB at 32K.
Where can I download Llama 3.2 3B Instruct Abliterated GGUF?
The files are on Hugging Face in the mradermacher/Llama-3.2-3B-Instruct-abliterated-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Llama 3.2 3B Instruct Abliterated use?
Hugging Face lists the license as llama3.2. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run this model on 8 GB VRAM?
Yes, Q4_K_M or Q5_K_S quantizations fit within 8 GB VRAM with 8k context.
Does this model support vision tasks?
No, the model has no vision capabilities.
What is the license of this model?
The model is licensed under llama3.2.