Dolphin3.0 Llama3.1 1B abliterated needs about 1.8 GB of memory at Q6_0 (a 1.0 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 15 quants, from 0.9 GB (Q4_0) to 2.5 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 1.1 GB.
Dolphin3.0 Llama3.1 1B abliterated – a 1.24B parameter Llama model
This model is a 1.24B parameter Llama architecture trained with 131072 context tokens. It is quantized by NexesQuants and released under the llama3.1 license. The model is available in multiple GGUF quantizations.
Good at
- Large context handling
- Low memory footprint
- Fast inference on compatible hardware
Limits
- No vision support
- Requires IK_Llama.cpp for inference
- License is llama3.1
Best for
- Chatbot development
- Rapid prototyping
- Embedded inference
Who it’s for. Developers and researchers needing a lightweight Llama model with extended context. The model is suitable for projects that require efficient inference on limited GPU resources.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 0.9 GB | 1.6 GB | Fits | Get file → | |
| Q5_0 | 0.9 GB | 1.7 GB | Fits | Get file → | |
| Q5_0 | 0.9 GB | 1.7 GB | Fits | Get file → | |
| Q5_K_S | 0.9 GB | 1.7 GB | Fits | Get file → | |
| Q5_K_S | 0.9 GB | 1.7 GB | Fits | Get file → | |
| Q4_1 | 0.9 GB | 1.7 GB | Fits | Get file → | |
| Q5_0 | 1.0 GB | 1.7 GB | Fits | Get file → | |
| Q6_0 | 1.0 GB | 1.8 GB | Fits | Get file → | |
| Q6_0 | 1.0 GB | 1.8 GB | Fits | Get file → | |
| Q6_K | 1.0 GB | 1.8 GB | Fits | Get file → | |
| Q6_K | 1.0 GB | 1.8 GB | Fits | Get file → | |
| Q5_1 | 1.0 GB | 1.8 GB | Fits | Get file → | |
| Q8_0 | 1.3 GB | 2.1 GB | Fits | Get file → | |
| BF16 | 2.5 GB | 3.2 GB | Fits | Get file → | |
| F16 | 2.5 GB | 3.2 GB | Fits | Get file → |
For 8 GB VRAM, Q4_0 (0.85 GB file, ~1.6 GB memory) is recommended. If 12 GB VRAM is available, Q5_K_S (0.89 GB file, ~1.7 GB memory) offers better quality. For 16 GB VRAM, Q8_0 (1.32 GB file, ~2.1 GB memory) provides the highest precision.
What GPU do you need for Dolphin3.0 Llama3.1 1B 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 BF16 (2.5 GB file, about 3.2 GB in use).
Which Dolphin3.0 Llama3.1 1B abliterated quant should you choose?
Q6_0 (1.0 GB) is the middle option in this repository. Very close to the original quality for most uses, at a clearly larger size than 4-bit.
If memory is tight, the smallest file is Q4_0 at 0.9 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Frequently asked questions about Dolphin3.0 Llama3.1 1B abliterated
How much VRAM does Dolphin3.0 Llama3.1 1B abliterated need?
About 1.8 GB at Q6_0 with an 8K context window. That is the 1.0 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 Dolphin3.0 Llama3.1 1B abliterated on an 8 GB GPU?
Yes, with BF16: the 2.5 GB file needs about 3.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Dolphin3.0 Llama3.1 1B abliterated GGUF?
It depends on the quant: Q4_0 is 0.9 GB, Q5_0 is 0.9 GB, Q5_0 is 0.9 GB, Q5_K_S is 0.9 GB, Q5_K_S is 0.9 GB, Q4_1 is 0.9 GB and 9 more.
What context length does Dolphin3.0 Llama3.1 1B 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 0.3 GB at 8K and 1.1 GB at 32K.
Where can I download Dolphin3.0 Llama3.1 1B abliterated GGUF?
The files are on Hugging Face in the NexesQuants/Dolphin3.0-Llama3.1-1B-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 Dolphin3.0 Llama3.1 1B abliterated use?
Hugging Face lists the license as llama3.1. 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, the Q4_0 quantization uses about 1.6 GB memory at 8 k context, fitting within 8 GB VRAM.
Does this model support vision tasks?
No, the model has no vision capabilities.
What inference engine is required?
Inference must use IK_Llama.cpp, not the standard llama.cpp.