BigWeave v28 96b needs about 62.0 GB of memory at Q4_K_M (a 57.8 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 18 quants, from 18.6 GB (Q5_K) to 66.3 GB (Q5_K_S).
BigWeave v28 96b is a 96.4B-parameter Bigweave model in GGUF format, quantized by RichardErkhov. Its trained context length is 32K tokens. The repository has 18 quant files, from 18.6 GB (Q5_K) to 66.3 GB (Q5_K_S).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q5_K | 18.6 GB | 22.8 GB | Too big | Get file → | |
| Q2_K | 35.4 GB | 39.7 GB | Too big | Get file → | |
| IQ3_XS | 39.4 GB | 43.7 GB | Too big | Get file → | |
| Q3_K_S | 41.5 GB (2 parts) | 45.8 GB | Too big | Get file → | |
| IQ3_S | 41.6 GB (2 parts) | 45.9 GB | Too big | Get file → | |
| IQ3_M | 43.1 GB (2 parts) | 47.3 GB | Too big | Get file → | |
| Q3_K | 46.3 GB (2 parts) | 50.6 GB | Too big | Get file → | |
| Q3_K_M | 46.3 GB (2 parts) | 50.6 GB | Too big | Get file → | |
| Q3_K_L | 50.5 GB (2 parts) | 54.7 GB | Too big | Get file → | |
| IQ4_XS | 51.9 GB (2 parts) | 56.2 GB | Too big | Get file → | |
| Q4_0 | 54.3 GB (2 parts) | 58.5 GB | Too big | Get file → | |
| Q4_K_S | 54.7 GB (2 parts) | 58.9 GB | Too big | Get file → | |
| IQ4_NL | 54.8 GB (2 parts) | 59.1 GB | Too big | Get file → | |
| Q4_K | 57.8 GB (2 parts) | 62.0 GB | Too big | Get file → | |
| Q4_K_M | 57.8 GB (2 parts) | 62.0 GB | Too big | Get file → | |
| Q4_1 | 60.3 GB (2 parts) | 64.5 GB | Too big | Get file → | |
| Q5_0 | 66.3 GB (2 parts) | 70.5 GB | Too big | Get file → | |
| Q5_K_S | 66.3 GB (2 parts) | 70.5 GB | Too big | Get file → |
What GPU do you need for BigWeave v28 96b?
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: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is Q5_K (18.6 GB file, about 22.8 GB in use).
- 32 GB VRAM: the best quant that fits is Q5_K (18.6 GB file, about 22.8 GB in use).
- 48 GB VRAM: the best quant that fits is IQ3_M (43.1 GB file, about 47.3 GB in use).
Which BigWeave v28 96b quant should you choose?
Q4_K_M is the usual starting point for BigWeave v28 96b: it balances file size and quality well. Here it is a 57.8 GB file.
If memory is tight, the smallest file is Q5_K at 18.6 GB, with the quality trade-offs that come with fewer bits. Closer to the original model than 4-bit, for a moderate increase in size. A good pick when you have memory to spare.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Frequently asked questions about BigWeave v28 96b
How much VRAM does BigWeave v28 96b need?
About 62.0 GB at Q4_K_M with an 8K context window. That is the 57.8 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 BigWeave v28 96b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 22.8 GB, so you would need a larger GPU or CPU offloading.
What is the file size of BigWeave v28 96b GGUF?
It depends on the quant: Q5_K is 18.6 GB, Q2_K is 35.4 GB, IQ3_XS is 39.4 GB, Q3_K_S is 41.5 GB, IQ3_S is 41.6 GB, IQ3_M is 43.1 GB and 12 more.
What context length does BigWeave v28 96b support?
The GGUF header lists a trained context length of 32,764 tokens (32K). The KV cache grows with the context you use: about 3.8 GB at 8K and 15.0 GB at 32K.
Where can I download BigWeave v28 96b GGUF?
The files are on Hugging Face in the RichardErkhov/llmixer_-_BigWeave-v28-96b-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 BigWeave v28 96b use?
Hugging Face lists the license as unknown. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.