Scal-lite-60b-code-math needs about 56.6 GB of memory at Q4_K_M (a 55.5 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 11 quants, from 41.9 GB (Q3_K_S) to 78.5 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 2.4 GB.
Scal-lite-60b-code-math is a 73.8B-parameter Scal model in GGUF format, quantized by mradermacher. Its trained context length is 128K tokens. The repository has 11 quant files, from 41.9 GB (Q3_K_S) to 78.5 GB (Q8_0).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q3_K_S | 41.9 GB | 43.0 GB | Too big | Get file → | |
| Q2_K | 41.9 GB | 43.0 GB | Too big | Get file → | |
| IQ4_XS | 42.4 GB | 43.5 GB | Too big | Get file → | |
| Q3_K_M | 45.0 GB | 46.1 GB | Too big | Get file → | |
| Q3_K_L | 46.5 GB | 47.6 GB | Too big | Get file → | |
| Q4_K_S | 51.2 GB | 52.3 GB | Too big | Get file → | |
| Q4_K_M | 55.5 GB | 56.6 GB | Too big | Get file → | |
| Q5_K_S | 55.6 GB | 56.7 GB | Too big | Get file → | |
| Q5_K_M | 59.4 GB | 60.5 GB | Too big | Get file → | |
| Q6_K | 78.4 GB | 79.5 GB | Too big | Get file → | |
| Q8_0 | 78.5 GB | 79.6 GB | Too big | Get file → |
What GPU do you need for Scal-lite-60b-code-math?
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: the best quant that fits is Q3_K_L (46.5 GB file, about 47.6 GB in use).
Which Scal-lite-60b-code-math quant should you choose?
Q4_K_M is the usual starting point for Scal-lite-60b-code-math: it balances file size and quality well. Here it is a 55.5 GB file.
The Q8_0 file is 78.5 GB, about 1.4 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q3_K_S at 41.9 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Scal-lite-60b-code-math
How much VRAM does Scal-lite-60b-code-math need?
About 56.6 GB at Q4_K_M with an 8K context window. That is the 55.5 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 Scal-lite-60b-code-math on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 43.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Scal-lite-60b-code-math GGUF?
It depends on the quant: Q3_K_S is 41.9 GB, Q2_K is 41.9 GB, IQ4_XS is 42.4 GB, Q3_K_M is 45.0 GB, Q3_K_L is 46.5 GB, Q4_K_S is 51.2 GB and 5 more.
What context length does Scal-lite-60b-code-math support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.6 GB at 8K and 2.4 GB at 32K.
Where can I download Scal-lite-60b-code-math GGUF?
The files are on Hugging Face in the mradermacher/scal-lite-60b-code-math-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 Scal-lite-60b-code-math 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.