From: <huggingface.co/mlx-community> Newsgroups: alt.freeweights.graniteSubject: granite-3.3-2b-instruct-4bit Date: 2026-11-15Commit: 58246c5Signed: 2DC28CC7B26A283B ✓ Text generation model with 2.5B parameters. Safetensors weights, 1.43 GB in all. Published by mlx-community under Apache-2.0.
2.5B params 1.43 GB granite-3.3-2b-instruct-4bit safetensors permissive Facts Params 2.5B Size 1.43 GB Formats safetensors Task text generation License Apache-2.0 permissive
fw
$ fw get mlx-community/granite-3.3-2b-instruct-4bit
# manifest sha256 bc83893be4f8… · 1.43 GB · 11 files
$ fw verify ./granite-3.3-2b-instruct-4bit mlx-community/granite-3.3-2b-instruct-4bitHosted (signed) mlx-community/granite-3.3-2b-instruct-4bit Last check: 2026-11-15 05:11 UTC
Get help: custom fine-tune or private deploy. We review your setup before we quote.
Quantized of granite-3.3-2b-instruct .
Works with
Runs with Ranked by fit to this model (safetensors), then by GitHub stars.
Show all 24 Fine-tune with Show all 29
Overview
Size 1.43 GB
Weights 1.43 GB (safetensors)
Parameters 2.5B
Formats safetensors
Task text generation
License Apache-2.0 (permissive)
Files 11, each with a sha256
Publisher mlx-community
Added 2026-11-15
Section alt.freeweights.granite Lineage
Original release from the publisher.
Base model: ibm-granite/granite-3.3-2b-instruct
Where these bytes come from
Publisher at the pinned commit →You
Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.
Sources
Badge Host Last check Result huggingface.co 2026-11-15 05:11 UTC passed
Model card From the publisher at commit
58246c5 . Shown as plain text with links; images open at the publisher.
This model mlx-community/granite-3.3-2b-instruct-4bit was converted to MLX format from ibm-granite/granite-3.3-2b-instruct using mlx-lm version 0.22.5 .
Use with mlx#
from mlx_lm import load, generate
model, tokenizer = load("mlx-community/granite-3.3-2b-instruct-4bit")
prompt = "hello"
if tokenizer.chat_template is not None:
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, add_generation_prompt=True
)
response = generate(model, tokenizer, prompt=prompt, verbose=True) Files 11 files · 1.43 GB Weights 1.43 GB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 11 files with their sha256 File Size sha256 .gitattributes 1.52 KB 11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361 README.md 956 B 9e4fa0f08ae1b0675d6372ea1a226798b46e0b4e14fd09797ec287d618fe68db added_tokens.json 207 B bb33d55934aa82d29cc62f3d19cdbc60f315763f6ccee21bdfd8b3bde2f33d3b config.json 1.00 KB ff28d67485d1a40022f0eb0c19b983c2680e90c25b690032e59e645f604f67d1 merges.txt 442 KB 303127a244b0078878156c17229f36d11b7a3a3f8e47b7cfdbb304ff46be5030 model.safetensors 1.43 GB f37ff14b14f2db07b4fc476e57e619b038b264fcc9f0681bc13274772698ac93 model.safetensors.index.json 65.3 KB 3fe2357794653e58654f56c0378e78e2a34bd6c9eac6b37563bc7896ccb54906 special_tokens_map.json 801 B 21ce694081bb9ae1bd4bc64549e72e0799ebb74705e6b650e3585d85b71ebdc1 tokenizer.json 3.48 MB 91168e938f05796aa6dcca7e485e4b30ab52785320c7a6391ecef86e6c84681e tokenizer_config.json 9.94 KB 042af89a5a0e0a183ebd104a875b1bf4435a091b19f2b62aaa58af1c5ea97600 vocab.json 777 KB 80ab859339a2525fdfbda14bc39df02dffb824aefdaf86426217bbb146d17e01
Versions Each time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commit Date Contents 58246c5498 pinned now 2026-11-15 05:11 UTC 11 files · 1.43 GB
Use it fw
$ fw get mlx-community/granite-3.3-2b-instruct-4bit
# manifest sha256 bc83893be4f8… · 1.43 GB · 11 files
$ fw verify ./granite-3.3-2b-instruct-4bit mlx-community/granite-3.3-2b-instruct-4bitfw.py Checks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/mlx-community/granite-3.3-2b-instruct-4bit.json --dir ./granite-3.3-2b-instruct-4bit
$ python3 fw.py verify ./granite-3.3-2b-instruct-4bit ./mlx-community__granite-3.3-2b-instruct-4bit.jsonhf The Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT =https://freeweights.allternit.com hf download mlx-community/granite-3.3-2b-instruct-4bit --revision 58246c5498495c14599525c852cfadb66c9f3084 --local-dir ./granite-3.3-2b-instruct-4bittransformers Load the weights in Python.
import os
os.environ["HF_ENDPOINT" ] = "https://freeweights.allternit.com"
from transformers import AutoModelForCausalLM, AutoTokenizer
tok = AutoTokenizer.from_pretrained("mlx-community/granite-3.3-2b-instruct-4bit", revision="58246c5498495c14599525c852cfadb66c9f3084")
model = AutoModelForCausalLM.from_pretrained("mlx-community/granite-3.3-2b-instruct-4bit", revision="58246c5498495c14599525c852cfadb66c9f3084", device_map="auto")vLLM Serve an OpenAI-compatible API.
$ HF_ENDPOINT =https://freeweights.allternit.com vllm serve mlx-community/granite-3.3-2b-instruct-4bit --revision 58246c5498495c14599525c852cfadb66c9f3084License Name Apache License 2.0 Id Apache-2.0 Class Permissive Redistribution yes Listing Hosted (signed) Decided License as declared by the publisher. No person read it. 2026-11-15 Text in README.md Text sha256 9e4fa0f08ae1b0675d6372ea1a226798b46e0b4e14fd09797ec287d618fe68db Terms gate No extra terms beyond the license. Read the license text
Licenses are checked against the license table . A custom license passes only after a person reads it.
Provenance From: <huggingface.co/mlx-community> Newsgroups: alt.freeweights.graniteSubject: granite-3.3-2b-instruct-4bit Date: 2026-11-15Commit: 58246c5Signed: 2DC28CC7B26A283B ✓ How to verify
$ curl -O https://freeweights.allternit.com/manifests/mlx-community/granite-3.3-2b-instruct-4bit.json
$ curl -O https://freeweights.allternit.com/manifests/mlx-community/granite-3.3-2b-instruct-4bit.json.minisig
$ minisign -Vm granite-3.3-2b-instruct-4bit.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq
$ shasum -a 256 <file> # compare with the Files tab A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download
Model tree No model in the catalog names this one as its parent.
Get it
$ HF_ENDPOINT =https://freeweights.allternit.com hf download mlx-community/granite-3.3-2b-instruct-4bitOpen in Freeweights Studio Pull, verify and run it in the Allternit app on Desktop, web or phone.
6.8B params 27.4 GB GneissWeb.7B_ablation_model_on_350B_FineWeb.Edu.seed1 safetensors permissive Facts Params 6.8B Size 27.4 GB Formats safetensors Task text generation License Apache-2.0 permissive flip From: ibm-granite <huggingface.co/ibm-granite> Subject: GneissWeb.7B_ablation_model_on_350B_FineWeb.Edu.seed1 6.8B params 27.4 GB GneissWeb.7B_ablation_model_on_350B_FineWeb.Edu.seed2 safetensors permissive Facts Params 6.8B Size 27.4 GB Formats safetensors Task text generation License Apache-2.0 permissive flip From: ibm-granite <huggingface.co/ibm-granite> Subject: GneissWeb.7B_ablation_model_on_350B_FineWeb.Edu.seed2 6.8B params 27.4 GB GneissWeb.7B_ablation_model_on_350B_FineWeb.Edu.seed3 safetensors permissive Facts Params 6.8B Size 27.4 GB Formats safetensors Task text generation License Apache-2.0 permissive flip From: ibm-granite <huggingface.co/ibm-granite> Subject: GneissWeb.7B_ablation_model_on_350B_FineWeb.Edu.seed3 6.8B params 27.4 GB GneissWeb.7B_ablation_model_on_350B_FineWeb.seed1 safetensors permissive Facts Params 6.8B Size 27.4 GB Formats safetensors Task text generation License Apache-2.0 permissive flip From: ibm-granite <huggingface.co/ibm-granite> Subject: GneissWeb.7B_ablation_model_on_350B_FineWeb.seed1 6.8B params 27.4 GB GneissWeb.7B_ablation_model_on_350B_FineWeb.seed2 safetensors permissive Facts Params 6.8B Size 27.4 GB Formats safetensors Task text generation License Apache-2.0 permissive flip From: ibm-granite <huggingface.co/ibm-granite> Subject: GneissWeb.7B_ablation_model_on_350B_FineWeb.seed2 6.8B params 27.4 GB GneissWeb.7B_ablation_model_on_350B_FineWeb.seed3 safetensors permissive Facts Params 6.8B Size 27.4 GB Formats safetensors Task text generation License Apache-2.0 permissive flip From: ibm-granite <huggingface.co/ibm-granite> Subject: GneissWeb.7B_ablation_model_on_350B_FineWeb.seed3
mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.graniteSubject: 
