Run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF locally
MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF is a large instruction-tuned chat model with 8.03 billion parameters, built on the llama architecture. It has been downloaded 177,782 times.
To run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF locally at a 4,096-token context, its quantized versions need between 3.18 GB (IQ1_S, lowest quality) and 16.27 GB (GGUF, highest quality) of memory, weights plus KV cache and a system margin included.
For most users the best balance is IQ2_XS, needing about 3.73 GB. That means MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF fits entirely in the VRAM of a 6 GB GPU or larger, running fully on the GPU.
All quantizations
| Quant. | Bits | Quality | Weights | KV | Total | Speed~ | Verdict |
|---|---|---|---|---|---|---|---|
| IQ1_S | 2.01 | Very low | 1.88 GB | 0.5 GB | 3.18 GB | 212.6 t/s | Fits in VRAM |
| IQ1_M | 2.15 | Very low | 2.01 GB | 0.5 GB | 3.31 GB | 198.6 t/s | Fits in VRAM |
| IQ2_XS | 2.6 | Very low | 2.43 GB | 0.5 GB | 3.73 GB | 164.8 t/s | Fits in VRAM |
| Q2_K | 3.17 | Low | 2.96 GB | 0.5 GB | 4.26 GB | 16.9 t/s | Offload |
| IQ3_XS | 3.51 | Fair | 3.28 GB | 0.5 GB | 4.58 GB | 15.3 t/s | Offload |
| Q3_K_S | 3.65 | Fair | 3.41 GB | 0.5 GB | 4.71 GB | 14.6 t/s | Offload |
| Q3_K_M | 4.0 | Fair | 3.74 GB | 0.5 GB | 5.04 GB | 13.4 t/s | Offload |
| Q3_K_L | 4.31 | Good | 4.03 GB | 0.5 GB | 5.33 GB | 12.4 t/s | Offload |
| IQ4_XS | 4.43 | Good | 4.14 GB | 0.5 GB | 5.44 GB | 12.1 t/s | Offload |
| Q4_K_S | 4.68 | Good | 4.37 GB | 0.5 GB | 5.67 GB | 11.4 t/s | Offload |
| Q4_K_M | 4.9 | Good | 4.58 GB | 0.5 GB | 5.88 GB | 10.9 t/s | Offload |
| Q5_K_S | 5.58 | Very good | 5.22 GB | 0.5 GB | 6.52 GB | 9.6 t/s | Offload |
| Q5_K_M | 5.71 | Very good | 5.34 GB | 0.5 GB | 6.64 GB | 9.4 t/s | Offload |
| Q6_K | 6.57 | Excellent | 6.14 GB | 0.5 GB | 7.44 GB | 8.1 t/s | Offload |
| Q8_0 | 8.51 | Excellent | 7.95 GB | 0.5 GB | 9.25 GB | 6.3 t/s | Offload |
| GGUF | 16.01 | Excellent | 14.97 GB | 0.5 GB | 16.27 GB | 3.3 t/s | Offload |
KV cache computed from the model's exact architecture. Speed is a rough estimate bounded by memory bandwidth.
Frequently asked questions
How much VRAM do you need to run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF?
You need about 5.88 GB of VRAM to run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF entirely on the GPU using the Q4_K_M quantization (at a 4,096-token context). Smaller quantizations lower the requirement at the cost of quality.
Can I run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF on an 8 GB GPU?
Yes. With 8 GB of VRAM you can run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF fully on the GPU using Q6_K (about 7.44 GB).
Can I run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF on a 16 GB GPU?
Yes. With 16 GB of VRAM you can run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF fully on the GPU using Q8_0 (about 9.25 GB).
Can I run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF on a 24 GB GPU?
Yes. With 24 GB of VRAM you can run MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF fully on the GPU using GGUF (about 16.27 GB).
What is the best quantization for MaziyarPanahi/Meta-Llama-3-8B-Instruct-GGUF?
If memory allows, higher bits-per-weight means better quality. A common sweet spot is a Q4_K_M or Q5_K_M quantization, which keeps most of the quality while roughly halving the memory versus 8-bit. Pick the highest quantization that still fits in your VRAM.