Instructions to use city96/t5-v1_1-xxl-encoder-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use city96/t5-v1_1-xxl-encoder-gguf with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="city96/t5-v1_1-xxl-encoder-gguf",
	filename="t5-v1_1-xxl-encoder-Q3_K_L.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use city96/t5-v1_1-xxl-encoder-gguf with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M

Use Docker

docker model run hf.co/city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M

LM Studio
Jan
Ollama
How to use city96/t5-v1_1-xxl-encoder-gguf with Ollama:
```
ollama run hf.co/city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M
```

Unsloth Studio new

How to use city96/t5-v1_1-xxl-encoder-gguf with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for city96/t5-v1_1-xxl-encoder-gguf to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for city96/t5-v1_1-xxl-encoder-gguf to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for city96/t5-v1_1-xxl-encoder-gguf to start chatting

Docker Model Runner
How to use city96/t5-v1_1-xxl-encoder-gguf with Docker Model Runner:
```
docker model run hf.co/city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M
```

Lemonade

How to use city96/t5-v1_1-xxl-encoder-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull city96/t5-v1_1-xxl-encoder-gguf:Q4_K_M

Run and chat with the model

lemonade run user.t5-v1_1-xxl-encoder-gguf-Q4_K_M

List all available models

lemonade list

Comparisons to FP8 e4m3fn ?

by NielsGx - opened Aug 20, 2024

Discussion

NielsGx

Aug 20, 2024

•

edited Aug 20, 2024

Hi,
I'm currently using the FP8 e4m3fn variant (4.55GB) on a 16GB VRAM GPU.

Mainly interested to know if Q6_K and Q8 would have better quality ?

And how would VRAM usage look ?
They both are relatively comparable in size (4GB and 5GB)

city96

Owner Aug 21, 2024

I think quality seems to be generally more faithful to FP16, especially with Q6_K and Q8_K. FP8 is a much simpler format. There's some comparison images in the comments of the PR where I added it though I don't think anyone tested this extensively. VRAM usage should be similar though comfy handles loading/unloading if it has to.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment