TO

Tot 360M Family

TotGGUF by verycoololiverUpdated 2026-11-01
Download on Hugging Face
Parameters345M
Context2K
Layers32
Architectureqwen3
Smallest quant0.3 GB
Quick answer

Tot 360M Family needs about 1.1 GB of memory at Q4_K_M (a 0.3 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 12 quants, from 0.3 GB (Q4_K_M) to 0.7 GB (F16).

Tot 360M Family – 345M‑parameter Qwen3 decoder-only model

Tot-360M Family is a 345,359,296‑parameter decoder‑only Transformer trained from scratch. It uses a 32,000‑token byte‑level BPE tokenizer and was trained on Kaggle Cloud TPU v5e‑8 and NVIDIA Tesla T4 GPUs. The model supports a 2,048‑token context window.

Good at

  • Byte‑level tokenization
  • Small‑parameter efficient inference
  • Community‑supported training pipeline

Limits

  • Context limited to 2,048 tokens
  • No vision or multimodal support
  • License not specified

Best for

  • Local chatbot deployments
  • Educational demos on Raspberry Pi
  • Low‑resource inference

Who it’s for. Developers and hobbyists looking for a lightweight, open‑source language model that can run on modest hardware. The model is suitable for local inference without requiring large GPU memory.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M0.3 GB 0.8 GB
Fits Get file →
Q4_K_M0.3 GB 0.8 GB
Fits Get file →
Q4_K_M0.3 GB 0.8 GB
Fits Get file →
Q4_K_M0.3 GB 0.8 GB
Fits Get file →
Q8_00.4 GB 1.0 GB
Fits Get file →
Q8_00.4 GB 1.0 GB
Fits Get file →
Q8_00.4 GB 1.0 GB
Fits Get file →
Q8_00.4 GB 1.0 GB
Fits Get file →
F160.7 GB 1.3 GB
Fits Get file →
F160.7 GB 1.3 GB
Fits Get file →
F160.7 GB 1.3 GB
Fits Get file →
F160.7 GB 1.3 GB
Fits Get file →

For 4 GB VRAM, use Q4_K_M (0.25 GB file, ~1.1 GB memory). For 8 GB VRAM, Q8_0 (0.37 GB file, ~1.2 GB memory) is a good balance. If you have more than 8 GB, F16 (0.69 GB file, ~1.5 GB memory) offers higher precision.

What GPU do you need for Tot 360M Family?

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: the best quant that fits is F16 (0.7 GB file, about 1.5 GB in use).

Which Tot 360M Family quant should you choose?

Q4_K_M is the usual starting point for Tot 360M Family: it balances file size and quality well. Here it is a 0.3 GB file.

The Q8_0 file is 0.4 GB, about 1.5 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.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen3
Parameters345M
Layers32
Trained context2,048 tokens
KV cache (fp16)40 KB per token

Source

Quantized byverycoololiver
FamilyTot
Downloads (HF)655
Last updated2026-11-01

Frequently asked questions about Tot 360M Family

How much VRAM does Tot 360M Family need?

About 1.1 GB at Q4_K_M with an 8K context window. That is the 0.3 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 Tot 360M Family on an 8 GB GPU?

Yes, with F16: the 0.7 GB file needs about 1.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Tot 360M Family GGUF?

It depends on the quant: Q4_K_M is 0.3 GB, Q4_K_M is 0.3 GB, Q4_K_M is 0.3 GB, Q4_K_M is 0.3 GB, Q8_0 is 0.4 GB, Q8_0 is 0.4 GB and 6 more.

What context length does Tot 360M Family support?

The GGUF header lists a trained context length of 2,048 tokens (2K). The KV cache grows with the context you use: about 0.3 GB at 8K and 0.1 GB at 2K.

Where can I download Tot 360M Family GGUF?

The files are on Hugging Face in the verycoololiver/Tot-360M-Family repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run Tot-360M on 8 GB VRAM?

Yes, the Q8_0 quantized model uses about 1.2 GB of VRAM at 8 k context, well within an 8 GB GPU.

What about running on a 4 GB Raspberry Pi 4?

The Q4_K_M quantized model requires roughly 1.1 GB of memory, which fits on a 4 GB Raspberry Pi 4 with some headroom.

Is there a version for 2 GB VRAM?

No quantized file is small enough for 2 GB VRAM; the smallest requires ~1.1 GB, so 2 GB would be tight but possible with careful memory management.

Similar models

Browse all →

Other tiny models

All tiny models →