TO

Torchcast Decision 31b

TorchcastGGUF by mradermacherUpdated 2026-10-28
Download on Hugging Face
Parameters32.1B
Context256K
Layers60
Architecturegemma4
Smallest quant12.4 GB
Quick answer

Torchcast Decision 31b needs about 52.2 GB of memory at Q4_K_M (a 19.5 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 11 quants, from 12.4 GB (Q2_K) to 34.1 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 128.8 GB.

Torchcast Decision 31b – a 32.1B Gemma‑4 decision model

Torchcast Decision 31b is a 32.1B parameter Gemma‑4 model fine‑tuned for decision tasks. It outputs yes/no probabilities, choice distributions, and ordinal score distributions without generating text. The model uses a hybrid attention scheme with global layers that attend bidirectionally.

Good at

  • Decision making
  • Yes/no probability output
  • Ordinal score prediction

Limits

  • No vision support
  • Requires large VRAM for higher quants
  • License is apache‑2.0

Best for

  • Automated decision systems
  • Chatbot decision engines
  • Score‑based recommendation tools

Who it’s for. Developers building decision‑oriented applications that need reliable probability outputs. The model is suitable for use cases where text generation is not required and a deterministic decision interface is preferred.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K12.4 GB 45.1 GB
Too big Get file →
Q3_K_S14.4 GB 47.1 GB
Too big Get file →
Q3_K_M15.9 GB 48.6 GB
Too big Get file →
Q3_K_L17.2 GB 49.9 GB
Too big Get file →
IQ4_XS17.6 GB 50.3 GB
Too big Get file →
Q4_K_S18.6 GB 51.3 GB
Too big Get file →
Q4_K_M19.5 GB 52.2 GB
Too big Get file →
Q5_K_S22.3 GB 55.0 GB
Too big Get file →
Q5_K_M22.8 GB 55.5 GB
Too big Get file →
Q6_K26.4 GB 59.1 GB
Too big Get file →
Q8_034.1 GB 66.8 GB
Too big Get file →

For 8 GB VRAM, use Q2_K (12.38 GB file, ~45.1 GB memory at 8k context). For 16 GB VRAM, Q4_K_M (19.48 GB file, ~52.2 GB memory) offers a balance of size and quality. For 32 GB VRAM, Q8_0 (34.13 GB file, ~66.8 GB memory) provides the highest fidelity.

What GPU do you need for Torchcast Decision 31b?

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_S (14.4 GB file, about 47.1 GB in use).

Which Torchcast Decision 31b quant should you choose?

Q4_K_M is the usual starting point for Torchcast Decision 31b: it balances file size and quality well. Here it is a 19.5 GB file.

The Q8_0 file is 34.1 GB, about 1.8 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 Q2_K at 12.4 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

Architecturegemma4
Parameters32.1B
Layers60
Trained context262,144 tokens
KV cache (fp16)3840 KB per token

Source

Quantized bymradermacher
FamilyTorchcast
Licenseapache-2.0
Downloads (HF)789
Last updated2026-10-28

Frequently asked questions about Torchcast Decision 31b

How much VRAM does Torchcast Decision 31b need?

About 52.2 GB at Q4_K_M with an 8K context window. That is the 19.5 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 Torchcast Decision 31b on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 45.1 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Torchcast Decision 31b GGUF?

It depends on the quant: Q2_K is 12.4 GB, Q3_K_S is 14.4 GB, Q3_K_M is 15.9 GB, Q3_K_L is 17.2 GB, IQ4_XS is 17.6 GB, Q4_K_S is 18.6 GB and 5 more.

What context length does Torchcast Decision 31b support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 32.2 GB at 8K and 128.8 GB at 32K.

Where can I download Torchcast Decision 31b GGUF?

The files are on Hugging Face in the mradermacher/torchcast-decision-31b-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 Torchcast Decision 31b 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.

Can I run Torchcast Decision 31b on 8 GB VRAM?

Use the Q2_K quant, which requires about 45.1 GB memory at 8k context. 8 GB VRAM is insufficient for any quant of this model.

Does this model support vision inputs?

No, the model is text‑only and does not include vision capabilities.

What license applies to Torchcast Decision 31b?

The model is released under the apache‑2.0 license.

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Torchcast models

All Torchcast →

Other medium models

All medium models →