TR

TriLM_3.9B_Unpacked

TrilmGGUF by Green-SkyUpdated 2024-08-03
Download on Hugging Face
Parameters3.99B
Context2K
Layers30
Architecturellama
Smallest quant1.0 GB
Quick answer

TriLM_3.9B_Unpacked needs about 5.8 GB of memory at Q4_0 (a 2.3 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 6 quants, from 1.0 GB (TQ1_0) to 8.0 GB (F16).

TriLM_3.9B_Unpacked is a 3.99B-parameter Trilm model in GGUF format, quantized by Green-Sky. Its trained context length is 2K tokens. The repository has 6 quant files, from 1.0 GB (TQ1_0) to 8.0 GB (F16).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
TQ1_01.0 GB 2.2 GB
Fits Get file →
TQ2_01.2 GB 2.4 GB
Fits Get file →
Q4_02.3 GB 3.5 GB
Fits Get file →
Q4_K2.4 GB 3.7 GB
Fits Get file →
Q8_04.2 GB 5.5 GB
Fits Get file →
F168.0 GB 9.2 GB
Too big Get file →

What GPU do you need for TriLM_3.9B_Unpacked?

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: the best quant that fits is Q4_K (2.4 GB file, about 5.9 GB in use).
  • 8 GB VRAM: the best quant that fits is Q8_0 (4.2 GB file, about 7.8 GB in use).
  • 12 GB VRAM: the best quant that fits is F16 (8.0 GB file, about 11.5 GB in use).

Which TriLM_3.9B_Unpacked quant should you choose?

Q4_0 (2.3 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.

If memory is tight, the smallest file is TQ1_0 at 1.0 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters3.99B
Layers30
Trained context2,048 tokens
KV cache (fp16)360 KB per token

Source

Quantized byGreen-Sky
FamilyTrilm
Licenseapache-2.0
Downloads (HF)131
Last updated2024-08-03

Frequently asked questions about TriLM_3.9B_Unpacked

How much VRAM does TriLM_3.9B_Unpacked need?

About 5.8 GB at Q4_0 with an 8K context window. That is the 2.3 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 TriLM_3.9B_Unpacked on an 8 GB GPU?

Yes, with Q8_0: the 4.2 GB file needs about 7.8 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of TriLM_3.9B_Unpacked GGUF?

It depends on the quant: TQ1_0 is 1.0 GB, TQ2_0 is 1.2 GB, Q4_0 is 2.3 GB, Q4_K is 2.4 GB, Q8_0 is 4.2 GB, F16 is 8.0 GB.

What context length does TriLM_3.9B_Unpacked support?

The GGUF header lists a trained context length of 2,048 tokens (2K). The KV cache grows with the context you use: about 3.0 GB at 8K and 0.8 GB at 2K.

Where can I download TriLM_3.9B_Unpacked GGUF?

The files are on Hugging Face in the Green-Sky/TriLM_3.9B-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 TriLM_3.9B_Unpacked 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.

Other GGUFs of TriLM_3.9B_Unpacked

Similar models

Browse all →

More from Green-Sky

All by Green-Sky →

More Trilm models

All Trilm →

Other small models

All small models →