SoftTiger 70b needs about 45.0 GB of memory at Q4_K_M (a 41.8 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 9 quants, from 25.8 GB (Q2_K) to 49.2 GB (Q5_K_M).
SoftTiger 70b is a 69.5B-parameter Softtiger model in GGUF format, quantized by mradermacher. Its trained context length is 8K tokens. The repository has 9 quant files, from 25.8 GB (Q2_K) to 49.2 GB (Q5_K_M).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 25.8 GB | 29.0 GB | Too big | Get file → | |
| Q3_K_S | 30.3 GB | 33.4 GB | Too big | Get file → | |
| Q3_K_M | 33.6 GB | 36.8 GB | Too big | Get file → | |
| Q3_K_L | 36.5 GB | 39.7 GB | Too big | Get file → | |
| IQ4_XS | 37.6 GB | 40.7 GB | Too big | Get file → | |
| Q4_K_S | 39.6 GB | 42.8 GB | Too big | Get file → | |
| Q4_K_M | 41.8 GB | 45.0 GB | Too big | Get file → | |
| Q5_K_S | 47.9 GB | 51.1 GB | Too big | Get file → | |
| Q5_K_M | 49.2 GB | 52.3 GB | Too big | Get file → |
What GPU do you need for SoftTiger 70b?
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 Q2_K (25.8 GB file, about 29.0 GB in use).
- 48 GB VRAM: the best quant that fits is Q4_K_M (41.8 GB file, about 45.0 GB in use).
Which SoftTiger 70b quant should you choose?
Q4_K_M is the usual starting point for SoftTiger 70b: it balances file size and quality well. Here it is a 41.8 GB file.
If memory is tight, the smallest file is Q2_K at 25.8 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 SoftTiger 70b
How much VRAM does SoftTiger 70b need?
About 45.0 GB at Q4_K_M with an 8K context window. That is the 41.8 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 SoftTiger 70b on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 29.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of SoftTiger 70b GGUF?
It depends on the quant: Q2_K is 25.8 GB, Q3_K_S is 30.3 GB, Q3_K_M is 33.6 GB, Q3_K_L is 36.5 GB, IQ4_XS is 37.6 GB, Q4_K_S is 39.6 GB and 3 more.
What context length does SoftTiger 70b support?
The GGUF header lists a trained context length of 8,192 tokens (8K). The KV cache grows with the context you use: about 2.7 GB at 8K and 2.7 GB at 8K.
Where can I download SoftTiger 70b GGUF?
The files are on Hugging Face in the mradermacher/SoftTiger-70b-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 SoftTiger 70b 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.