You can deploy this model and start earning money today!

REALMS-CG/
granite3.3-8b-20260526

Name: granite3.3-8b-20260526
Availability: InStock

gguf

endpoints_compatible

conversational

Views

No views yet

granite3.3-8b-20260526

Frozen GGUF export of Ollama model:

granite3.3:8b

Hugging Face repository:

REALMS-CG/granite3.3-8b-20260526

Files

granite3.3-8b-20260526.gguf
SHA256SUMS.txt
Modelfile
Original.Ollama.Modelfile.txt
Original.Ollama.Show.txt

Download from Hugging Face

Install the Hugging Face CLI:

bash

python3 -m venv .venv
source .venv/bin/activate
python3 -m pip install -U huggingface_hub

Download the full repository:

bash

hf download REALMS-CG/granite3.3-8b-20260526 \
  --repo-type model \
  --local-dir ./granite3.3-8b-20260526

Enter the downloaded folder:

cd granite3.3-8b-20260526

Verify checksum

sha256sum -c SHA256SUMS.txt

Expected output:

granite3.3-8b-20260526.gguf: OK

Recreate in local Ollama

ollama create granite3.3-8b-20260526 -f Modelfile

Run in local Ollama

ollama run granite3.3-8b-20260526

Recreate in Docker Ollama

If Ollama is running inside Docker with container name local-ollama-gpu, run:

bash

docker cp ./granite3.3-8b-20260526 local-ollama-gpu:/models/granite3.3-8b-20260526

docker exec -it local-ollama-gpu ollama create granite3.3-8b-20260526 \
  -f /models/granite3.3-8b-20260526/Modelfile

Run the model:

docker exec -it local-ollama-gpu ollama run granite3.3-8b-20260526

Reproducibility

This model should only be considered byte-for-byte reproducible if checksum verification succeeds:

sha256sum -c SHA256SUMS.txt

Expected:

granite3.3-8b-20260526.gguf: OK

Original Ollama model information

The original Ollama metadata is stored in:

Original.Ollama.Modelfile.txt
Original.Ollama.Show.txt