thinkingmachines.Inkling Small needs about 165.5 GB of memory at F16 (a 159.3 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 6 quants, from 96.0 GB (F16) to 280.3 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 22.5 GB.
thinkingmachines.Inkling Small is a 264B-parameter Inkling model in GGUF format, quantized by DevQuasar. Its trained context length is 1024K tokens. The repository has 6 quant files, from 96.0 GB (F16) to 280.3 GB (F16).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| F16 | 96.0 GB (8 parts) | 102.2 GB | Too big | Get file → | |
| F16 | 125.4 GB (10 parts) | 131.6 GB | Too big | Get file → | |
| F16 | 159.3 GB (13 parts) | 165.5 GB | Too big | Get file → | |
| F16 | 187.0 GB (14 parts) | 193.2 GB | Too big | Get file → | |
| F16 | 216.4 GB (18 parts) | 222.6 GB | Too big | Get file → | |
| F16 | 280.3 GB (21 parts) | 286.4 GB | Too big | Get file → |
What GPU do you need for thinkingmachines.Inkling Small?
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: no quant fits at an 8K context.
- 48 GB VRAM: no quant fits at an 8K context.
Which thinkingmachines.Inkling Small quant should you choose?
F16 (159.3 GB) is the middle option in this repository. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
If memory is tight, the smallest file is F16 at 96.0 GB, with the quality trade-offs that come with fewer bits. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about thinkingmachines.Inkling Small
How much VRAM does thinkingmachines.Inkling Small need?
About 165.5 GB at F16 with an 8K context window. That is the 159.3 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 thinkingmachines.Inkling Small on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 102.2 GB, so you would need a larger GPU or CPU offloading.
What is the file size of thinkingmachines.Inkling Small GGUF?
It depends on the quant: F16 is 96.0 GB, F16 is 125.4 GB, F16 is 159.3 GB, F16 is 187.0 GB, F16 is 216.4 GB, F16 is 280.3 GB.
What context length does thinkingmachines.Inkling Small support?
The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 5.6 GB at 8K and 22.5 GB at 32K.
Where can I download thinkingmachines.Inkling Small GGUF?
The files are on Hugging Face in the DevQuasar/thinkingmachines.Inkling-Small-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 thinkingmachines.Inkling Small use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.