Yi-Coder-9B-Chat GGUF size and VRAM requirements
MaziyarPanahi/Yi-Coder-9B-Chat-GGUF is a large code-focused language model with 8.83 billion parameters, built on the llama architecture. It has been downloaded 141,076 times.
Under the hood it uses 48 transformer layers, a hidden size of 4,096, 32 attention heads. It uses grouped-query attention (32 query heads sharing 4 key/value heads), which already trims KV-cache memory compared with full multi-head attention.
To run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF locally at a 4,096-token context, its quantized versions need between 3.05 GB (IQ1_S, lowest quality) and 17.62 GB (GGUF, highest quality) of memory, weights plus KV cache and a system margin included. Context length drives that memory directly: at 4,096 tokens the KV cache for Yi-Coder-9B-Chat-GGUF is about 0.38 GB, rising to roughly 12.0 GB at its full 131,072-token context. Shorter prompts free up memory for a higher-quality quantization.
For most users the best balance is Q5_K_M, needing about 7.0 GB. That means MaziyarPanahi/Yi-Coder-9B-Chat-GGUF fits entirely in the VRAM of a 6 GB GPU or larger, running fully on the GPU.
Available GGUF quantizations for MaziyarPanahi/Yi-Coder-9B-Chat-GGUF include IQ1_S, IQ1_M, IQ2_XS, Q2_K, IQ3_XS, Q3_K_S, Q3_K_M, Q3_K_L, IQ4_XS, Q4_K_S, Q4_K_M, Q5_K_S, Q5_K_M, GGUF. The model supports a native context length of up to 131,072 tokens; a longer context grows the KV cache and the memory needed.
GGUF file size and memory by quantization
Compare real GGUF weight sizes, estimated KV cache and total memory for Q4, Q5, Q8 and every quantization published in this repository.
| Quant. | Bits | Quality | Weights | KV | Total | Speed~ | Verdict |
|---|---|---|---|---|---|---|---|
| IQ1_S | 1.83 | Very low | 1.88 GB | 0.38 GB | 3.05 GB | 213.2 t/s | Fits in VRAM |
| IQ1_M | 1.98 | Very low | 2.03 GB | 0.38 GB | 3.21 GB | 196.9 t/s | Fits in VRAM |
| IQ2_XS | 2.45 | Very low | 2.52 GB | 0.38 GB | 3.7 GB | 158.6 t/s | Fits in VRAM |
| Q2_K | 3.04 | Low | 3.12 GB | 0.38 GB | 4.3 GB | 128.0 t/s | Fits in VRAM |
| IQ3_XS | 3.37 | Fair | 3.46 GB | 0.38 GB | 4.64 GB | 115.5 t/s | Fits in VRAM |
| Q3_K_S | 3.53 | Fair | 3.63 GB | 0.38 GB | 4.81 GB | 110.1 t/s | Fits in VRAM |
| Q3_K_M | 3.92 | Fair | 4.03 GB | 0.38 GB | 5.2 GB | 99.3 t/s | Fits in VRAM |
| Q3_K_L | 4.25 | Good | 4.37 GB | 0.38 GB | 5.54 GB | 91.6 t/s | Fits in VRAM |
| IQ4_XS | 4.34 | Good | 4.46 GB | 0.38 GB | 5.63 GB | 89.8 t/s | Fits in VRAM |
| Q4_K_S | 4.6 | Good | 4.72 GB | 0.38 GB | 5.9 GB | 84.7 t/s | Fits in VRAM |
| Q4_K_M | 4.83 | Good | 4.96 GB | 0.38 GB | 6.14 GB | 80.6 t/s | Fits in VRAM |
| Q5_K_S | 5.53 | Very good | 5.69 GB | 0.38 GB | 6.86 GB | 70.3 t/s | Fits in VRAM |
| Q5_K_M | 5.67 | Very good | 5.83 GB | 0.38 GB | 7.0 GB | 68.6 t/s | Fits in VRAM |
| GGUF | 16.0 | Excellent | 16.45 GB | 0.38 GB | 17.62 GB | 3.0 t/s | Offload |
KV cache computed from the model's exact architecture. Speed is a rough estimate bounded by memory bandwidth.
Frequently asked questions
What kind of model is MaziyarPanahi/Yi-Coder-9B-Chat-GGUF?
MaziyarPanahi/Yi-Coder-9B-Chat-GGUF is a code-focused language model with 8.83 billion parameters, based on the llama architecture. It is distributed as GGUF files for local inference.
How much VRAM do you need to run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF?
You need about 5.9 GB of VRAM to run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF entirely on the GPU using the Q4_K_S quantization (at a 4,096-token context). Smaller quantizations lower the requirement at the cost of quality.
Can I run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF on an 8 GB GPU?
Yes. With 8 GB of VRAM you can run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF fully on the GPU using Q5_K_M (about 7.0 GB).
Can I run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF on a 16 GB GPU?
Yes. With 16 GB of VRAM you can run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF fully on the GPU using Q5_K_M (about 7.0 GB).
Can I run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF on a 24 GB GPU?
Yes. With 24 GB of VRAM you can run MaziyarPanahi/Yi-Coder-9B-Chat-GGUF fully on the GPU using GGUF (about 17.62 GB).
What context length does MaziyarPanahi/Yi-Coder-9B-Chat-GGUF support?
MaziyarPanahi/Yi-Coder-9B-Chat-GGUF supports a native context length of up to 131,072 tokens. A longer context grows the KV cache, so it increases the memory needed to run the model.
What is the best quantization for MaziyarPanahi/Yi-Coder-9B-Chat-GGUF?
For MaziyarPanahi/Yi-Coder-9B-Chat-GGUF, a strong default is Q4_K_M, which needs about 6.14 GB and keeps most of the quality while roughly halving the memory versus 8-bit. With VRAM to spare, Q5_K_M or Q6_K add a little more quality; if you are tight on memory, a smaller quantization still runs. Pick the highest quantization that fits your VRAM.