From: <huggingface.co/mlx-community> Newsgroups: alt.freeweights.olmoSubject: Olmo-3-7B-Instruct-mxfp4-QAT Date: 2026-11-15Commit: ad7ea40Signed: 2DC28CC7B26A283B ✓ Text generation model with 7.3B parameters. Safetensors weights, 3.89 GB in all. Published by mlx-community under Apache-2.0.
7.3B params 3.89 GB Olmo-3-7B-Instruct-mxfp4-QAT safetensors permissive Facts Params 7.3B Size 3.89 GB Formats safetensors Task text generation License Apache-2.0 permissive
fw
$ fw get mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT
# manifest sha256 1f574107031e… · 3.89 GB · 8 files
$ fw verify ./Olmo-3-7B-Instruct-mxfp4-QAT mlx-community/Olmo-3-7B-Instruct-mxfp4-QATHosted (signed) mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT Last check: 2026-11-15 05:15 UTC
Get help: custom fine-tune or private deploy. We review your setup before we quote.
Quantized of Olmo-3-7B-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 3.89 GB
Weights 3.88 GB (safetensors)
Parameters 7.3B
Formats safetensors
Task text generation
License Apache-2.0 (permissive)
Files 8, each with a sha256
Publisher mlx-community
Added 2026-11-15
Section alt.freeweights.olmo Lineage
Original release from the publisher.
Base model: allenai/Olmo-3-7B-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 05:15 UTC passed
Model card From the publisher at commit
ad7ea40 . Shown as plain text with links; images open at the publisher.
MLX-LM-LoRA Model#
This model was fine-tuned using mlx-lm-lora version 2.1.0.
Model Details#
Base model: allenai/Olmo-3-7B-Instruct Model type: olmo3 Training method: QAT + LoRA fine-tuning with MLX Fusion date: 2026-04-22 23:46:40 Usage#
This model can be loaded and used with the MLX framework.
Command to recreate#
python -m mlx_lm_lora.train \
--model allenai/Olmo-3-7B-Instruct \
--train-type lora \
--optimizer adamw \
--steps-per-report 1 \
--iters 100 \
--max-seq-length 4096 \
--adapter-path ./Olmo-3-7B-Instruct-mxfp4-QAT \
--data mlx-community/Dolci-Instruct-SFT-No-Tools-400K \
--train \
--qat-enable \
--qat-bits 4 \
--qat-group-size 64 \
--qat-mode mxfp4 \
--qat-start-step 1 \
--qat-interval 1Files 8 files · 3.89 GB Weights 3.88 GB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 8 files with their sha256 File Size sha256 .gitattributes 1.52 KB 11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361 README.md 971 B 12c08f221c5acf4e0f2cbcce2b459931f962f00948e748a5af90a3d0de70eb9a chat_template.jinja 2.61 KB f5186d42d99c8a0445d37fd8a6c7ccf07fe3e24a29ce622d8bd245da9507b12b config.json 1.77 KB 420cc0d7dd9e3344127c287429afe41119ba35e248c2ffe34cd337e13e421887 model.safetensors 3.88 GB 8e164a5161cb0ffd871e5c5284076d06a44b95a610edf29cc788e9a345f0569f model.safetensors.index.json 41.5 KB 73b3235a9fc83330b632acd56d5a1794cb3e8dd8af26dfc33e5142ebd0713459 tokenizer.json 7.14 MB 66becb9a96db60fb124531b6b12d911ca9c69e4f35d453f6f283dbdfe9c013d4 tokenizer_config.json 332 B 992c9ebe9f9ba5206ec1b71e220622028bead98ae6dbb645e192d2e642f959fd
Versions Each time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commit Date Contents ad7ea408b6 pinned now 2026-11-15 05:15 UTC 8 files · 3.89 GB
Use it fw
$ fw get mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT
# manifest sha256 1f574107031e… · 3.89 GB · 8 files
$ fw verify ./Olmo-3-7B-Instruct-mxfp4-QAT mlx-community/Olmo-3-7B-Instruct-mxfp4-QATfw.py Checks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT.json --dir ./Olmo-3-7B-Instruct-mxfp4-QAT
$ python3 fw.py verify ./Olmo-3-7B-Instruct-mxfp4-QAT ./mlx-community__Olmo-3-7B-Instruct-mxfp4-QAT.jsonhf The Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT =https://freeweights.allternit.com hf download mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT --revision ad7ea408b60ecb814dcc9a0f408f4e27ed1a4b75 --local-dir ./Olmo-3-7B-Instruct-mxfp4-QATtransformers 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/Olmo-3-7B-Instruct-mxfp4-QAT", revision="ad7ea408b60ecb814dcc9a0f408f4e27ed1a4b75")
model = AutoModelForCausalLM.from_pretrained("mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT", revision="ad7ea408b60ecb814dcc9a0f408f4e27ed1a4b75", device_map="auto")vLLM Serve an OpenAI-compatible API.
$ HF_ENDPOINT =https://freeweights.allternit.com vllm serve mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT --revision ad7ea408b60ecb814dcc9a0f408f4e27ed1a4b75License Name Apache License 2.0 Id Apache-2.0 Class Permissive Redistribution yes Listing Hosted (signed) Decided Inherited from the base model allenai/Olmo-3-7B-Instruct: the strictest license along the lineage applies. 2026-11-15 Text in README.md Text sha256 12c08f221c5acf4e0f2cbcce2b459931f962f00948e748a5af90a3d0de70eb9a 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.olmoSubject: Olmo-3-7B-Instruct-mxfp4-QAT Date: 2026-11-15Commit: ad7ea40Signed: 2DC28CC7B26A283B ✓ How to verify
$ curl -O https://freeweights.allternit.com/manifests/mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT.json
$ curl -O https://freeweights.allternit.com/manifests/mlx-community/Olmo-3-7B-Instruct-mxfp4-QAT.json.minisig
$ minisign -Vm Olmo-3-7B-Instruct-mxfp4-QAT.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/Olmo-3-7B-Instruct-mxfp4-QATOpen in Freeweights Studio Pull, verify and run it in the Allternit app on Desktop, web or phone.
olmo 114 GB allenai_Olmo-3-7B-Instruct-GGUF gguf permissive Facts Params not stated Size 114 GB Formats gguf Task text generation License Apache-2.0 permissive flip From: bartowski <huggingface.co/bartowski> Subject: allenai_Olmo-3-7B-Instruct-GGUF olmo 114 GB allenai_Olmo-3-7B-Think-GGUF gguf permissive Facts Params not stated Size 114 GB Formats gguf Task text generation License Apache-2.0 permissive flip From: bartowski <huggingface.co/bartowski> Subject: allenai_Olmo-3-7B-Think-GGUF olmo 520 GB allenai_Olmo-3.1-32B-Instruct-GGUF gguf permissive Facts Params not stated Size 520 GB Formats gguf Task text generation License Apache-2.0 permissive flip From: bartowski <huggingface.co/bartowski> Subject: allenai_Olmo-3.1-32B-Instruct-GGUF olmo 520 GB allenai_Olmo-3.1-32B-Think-GGUF gguf permissive Facts Params not stated Size 520 GB Formats gguf Task text generation License Apache-2.0 permissive flip From: bartowski <huggingface.co/bartowski> Subject: allenai_Olmo-3.1-32B-Think-GGUF olmo 123 GB allenai_olmOCR-2-7B-1025-GGUF gguf permissive Facts Params not stated Size 123 GB Formats gguf Task image text to text License Apache-2.0 permissive flip From: bartowski <huggingface.co/bartowski> Subject: allenai_olmOCR-2-7B-1025-GGUF olmo 106 GB allenai_olmOCR-7B-0225-preview-GGUF gguf permissive Facts Params not stated Size 106 GB Formats gguf Task image text to text License Apache-2.0 permissive flip From: bartowski <huggingface.co/bartowski> Subject: allenai_olmOCR-7B-0225-preview-GGUF
mlx-community <huggingface.co/mlx-community>Newsgroups: alt.freeweights.olmoSubject: 
