Babel 83B needs about 39.9 GB of memory at IQ3_XXS (a 36.3 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 17 quants, from 25.9 GB (IQ1_S) to 47.3 GB (Q4_0). Longer context needs more: at 32K tokens the KV cache alone adds about 12.3 GB.
Babel 83B is a 83.2B-parameter Babel model in GGUF format, quantized by mradermacher. Its trained context length is 128K tokens. The repository has 17 quant files, from 25.9 GB (IQ1_S) to 47.3 GB (Q4_0).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ1_S | 25.9 GB | 29.5 GB | Too big | Get file → | |
| IQ1_M | 27.1 GB | 30.7 GB | Too big | Get file → | |
| IQ2_XXS | 29.1 GB | 32.7 GB | Too big | Get file → | |
| IQ2_XS | 30.9 GB | 34.5 GB | Too big | Get file → | |
| IQ2_S | 31.9 GB | 35.5 GB | Too big | Get file → | |
| IQ2_M | 33.5 GB | 37.1 GB | Too big | Get file → | |
| Q2_K_S | 33.7 GB | 37.3 GB | Too big | Get file → | |
| Q2_K | 34.0 GB | 37.6 GB | Too big | Get file → | |
| IQ3_XXS | 36.3 GB | 39.9 GB | Too big | Get file → | |
| IQ3_XS | 37.4 GB | 41.0 GB | Too big | Get file → | |
| Q3_K_S | 39.2 GB | 42.8 GB | Too big | Get file → | |
| IQ3_S | 39.3 GB | 42.9 GB | Too big | Get file → | |
| IQ3_M | 40.5 GB | 44.1 GB | Too big | Get file → | |
| Q3_K_M | 43.0 GB | 46.6 GB | Too big | Get file → | |
| Q3_K_L | 45.2 GB | 48.8 GB | Too big | Get file → | |
| IQ4_XS | 45.4 GB | 49.0 GB | Too big | Get file → | |
| Q4_0 | 47.3 GB | 50.9 GB | Too big | Get file → |
What GPU do you need for Babel 83B?
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: no quant fits at an 8K context.
- 32 GB VRAM: the best quant that fits is IQ1_M (27.1 GB file, about 30.7 GB in use).
- 48 GB VRAM: the best quant that fits is Q3_K_M (43.0 GB file, about 46.6 GB in use).
Which Babel 83B quant should you choose?
IQ3_XXS (36.3 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
If memory is tight, the smallest file is IQ1_S at 25.9 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
Source
Frequently asked questions about Babel 83B
How much VRAM does Babel 83B need?
About 39.9 GB at IQ3_XXS with an 8K context window. That is the 36.3 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 Babel 83B on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 29.5 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Babel 83B GGUF?
It depends on the quant: IQ1_S is 25.9 GB, IQ1_M is 27.1 GB, IQ2_XXS is 29.1 GB, IQ2_XS is 30.9 GB, IQ2_S is 31.9 GB, IQ2_M is 33.5 GB and 11 more.
What context length does Babel 83B support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 3.1 GB at 8K and 12.3 GB at 32K.
Where can I download Babel 83B GGUF?
The files are on Hugging Face in the mradermacher/Babel-83B-i1-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 Babel 83B use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.