Instructions to use tensorblock/granite-3.0-3b-a800m-base-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="tensorblock/granite-3.0-3b-a800m-base-GGUF")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("tensorblock/granite-3.0-3b-a800m-base-GGUF", dtype="auto")

llama-cpp-python

How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="tensorblock/granite-3.0-3b-a800m-base-GGUF",
	filename="granite-3.0-3b-a800m-base-Q2_K.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K
# Run inference directly in the terminal:
llama cli -hf tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K
# Run inference directly in the terminal:
llama cli -hf tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K

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 tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K
# Run inference directly in the terminal:
./llama-cli -hf tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K

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 tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K
# Run inference directly in the terminal:
./build/bin/llama-cli -hf tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K

Use Docker

docker model run hf.co/tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K

LM Studio
Jan

vLLM

How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tensorblock/granite-3.0-3b-a800m-base-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tensorblock/granite-3.0-3b-a800m-base-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K

SGLang

How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "tensorblock/granite-3.0-3b-a800m-base-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tensorblock/granite-3.0-3b-a800m-base-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "tensorblock/granite-3.0-3b-a800m-base-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tensorblock/granite-3.0-3b-a800m-base-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Ollama
How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with Ollama:
```
ollama run hf.co/tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K
```

Unsloth Studio

How to use tensorblock/granite-3.0-3b-a800m-base-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 tensorblock/granite-3.0-3b-a800m-base-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 tensorblock/granite-3.0-3b-a800m-base-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for tensorblock/granite-3.0-3b-a800m-base-GGUF to start chatting

Atomic Chat new
Docker Model Runner
How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with Docker Model Runner:
```
docker model run hf.co/tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K
```

Lemonade

How to use tensorblock/granite-3.0-3b-a800m-base-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull tensorblock/granite-3.0-3b-a800m-base-GGUF:Q2_K

Run and chat with the model

lemonade run user.granite-3.0-3b-a800m-base-GGUF-Q2_K

List all available models

lemonade list

granite-3.0-3b-a800m-base-GGUF / README.md

morriszms

Update README.md

add02c9 verified 12 months ago

preview code

Raw

History Blame Contribute Delete

8.7 kB

metadata

pipeline_tag: text-generation
inference: false
license: apache-2.0
library_name: transformers
tags:
  - language
  - granite-3.0
  - TensorBlock
  - GGUF
base_model: ibm-granite/granite-3.0-3b-a800m-base
model-index:
  - name: granite-3.0-3b-a800m-base
    results:
      - task:
          type: text-generation
        dataset:
          name: MMLU
          type: human-exams
        metrics:
          - type: pass@1
            value: 48.64
            name: pass@1
          - type: pass@1
            value: 18.84
            name: pass@1
          - type: pass@1
            value: 23.81
            name: pass@1
      - task:
          type: text-generation
        dataset:
          name: WinoGrande
          type: commonsense
        metrics:
          - type: pass@1
            value: 65.67
            name: pass@1
          - type: pass@1
            value: 42.2
            name: pass@1
          - type: pass@1
            value: 47.39
            name: pass@1
          - type: pass@1
            value: 78.29
            name: pass@1
          - type: pass@1
            value: 72.79
            name: pass@1
          - type: pass@1
            value: 41.34
            name: pass@1
      - task:
          type: text-generation
        dataset:
          name: BoolQ
          type: reading-comprehension
        metrics:
          - type: pass@1
            value: 75.75
            name: pass@1
          - type: pass@1
            value: 20.96
            name: pass@1
      - task:
          type: text-generation
        dataset:
          name: ARC-C
          type: reasoning
        metrics:
          - type: pass@1
            value: 46.84
            name: pass@1
          - type: pass@1
            value: 24.83
            name: pass@1
          - type: pass@1
            value: 38.93
            name: pass@1
          - type: pass@1
            value: 35.05
            name: pass@1
      - task:
          type: text-generation
        dataset:
          name: HumanEval
          type: code
        metrics:
          - type: pass@1
            value: 26.83
            name: pass@1
          - type: pass@1
            value: 34.6
            name: pass@1
      - task:
          type: text-generation
        dataset:
          name: GSM8K
          type: math
        metrics:
          - type: pass@1
            value: 35.86
            name: pass@1
          - type: pass@1
            value: 17.4
            name: pass@1

ibm-granite/granite-3.0-3b-a800m-base - GGUF

This repo contains GGUF format model files for ibm-granite/granite-3.0-3b-a800m-base.

The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b4011.

Our projects

Forge

An OpenAI-compatible multi-provider routing layer.
🚀 Try it now! 🚀
Awesome MCP Servers	TensorBlock Studio

A comprehensive collection of Model Context Protocol (MCP) servers.	A lightweight, open, and extensible multi-LLM interaction studio.
👀 See what we built 👀	👀 See what we built 👀

## Prompt template

Model file specification

Filename	Quant type	File Size	Description
granite-3.0-3b-a800m-base-Q2_K.gguf	Q2_K	1.266 GB	smallest, significant quality loss - not recommended for most purposes
granite-3.0-3b-a800m-base-Q3_K_S.gguf	Q3_K_S	1.488 GB	very small, high quality loss
granite-3.0-3b-a800m-base-Q3_K_M.gguf	Q3_K_M	1.644 GB	very small, high quality loss
granite-3.0-3b-a800m-base-Q3_K_L.gguf	Q3_K_L	1.774 GB	small, substantial quality loss
granite-3.0-3b-a800m-base-Q4_0.gguf	Q4_0	1.926 GB	legacy; small, very high quality loss - prefer using Q3_K_M
granite-3.0-3b-a800m-base-Q4_K_S.gguf	Q4_K_S	1.942 GB	small, greater quality loss
granite-3.0-3b-a800m-base-Q4_K_M.gguf	Q4_K_M	2.059 GB	medium, balanced quality - recommended
granite-3.0-3b-a800m-base-Q5_0.gguf	Q5_0	2.338 GB	legacy; medium, balanced quality - prefer using Q4_K_M
granite-3.0-3b-a800m-base-Q5_K_S.gguf	Q5_K_S	2.338 GB	large, low quality loss - recommended
granite-3.0-3b-a800m-base-Q5_K_M.gguf	Q5_K_M	2.407 GB	large, very low quality loss - recommended
granite-3.0-3b-a800m-base-Q6_K.gguf	Q6_K	2.776 GB	very large, extremely low quality loss
granite-3.0-3b-a800m-base-Q8_0.gguf	Q8_0	3.593 GB	very large, extremely low quality loss - not recommended

Downloading instruction

Command line

Firstly, install Huggingface Client

pip install -U "huggingface_hub[cli]"

Then, downoad the individual model file the a local directory

huggingface-cli download tensorblock/granite-3.0-3b-a800m-base-GGUF --include "granite-3.0-3b-a800m-base-Q2_K.gguf" --local-dir MY_LOCAL_DIR

If you wanna download multiple model files with a pattern (e.g., *Q4_K*gguf), you can try:

huggingface-cli download tensorblock/granite-3.0-3b-a800m-base-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'