# RedHatAI/Llama-3.2-3B-Instruct-FP8-dynamic

This model is a quantized version of Llama 3.2 3B Instruct with FP8 weights and FP8 activations, and the card says it was obtained by quantizing the weights and activations to FP8. It is intended for commercial and research use in multiple languages, and it is suited to assistant like chat.

## Cite this

- Page: https://huggingbay.xyz/artifact/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic
- Citation: RedHatAI/Llama-3.2-3B-Instruct-FP8-dynamic. The Hugging Bay model record. https://huggingbay.xyz/artifact/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic (retrieved 2026-11-14).
- Check a downloaded file: run `shasum -a 256 <file>` (macOS), `sha256sum <file>` (Linux) or `Get-FileHash -Algorithm SHA256 <file>` (Windows), then compare with that file's SHA-256 at https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/files?limit=100. How checks work: https://huggingbay.xyz/verification

## Facts

- Type: LLM
- Task: text-generation
- Framework: unknown
- License: llama3.2: Commercial use is permitted with obligations: Llama community license: commercial use allowed; if your products/services exceeded 700 million monthly active users (Section 2, Additional Commercial Terms) you must request a license from Meta first; the acceptable-use policy and attribution notice apply.
- Source: https://huggingface.co/RedHatAI/Llama-3.2-3B-Instruct-FP8-dynamic
- Hosting: hosted (6 files)
- Files: 6 hosted of 9 listed
- File counts: "listed" is every file in the source repository; only hosted files download from Hugging Bay, and the download plan lists the files it selects for one download, so its count can differ.
- Hosted here: 6 hosted of 9 listed
- Full repo upstream: 4.41 GB
- Recommended file: 4.4 GB · model.safetensors
- Parameters: 3B
- Download: https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/download-plan (download plan: https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/download-plan)
- Per-file SHA-256: available for 6 of 6 hosted files via https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/files?limit=100
- Availability: Ready
- Downloadable: yes
- Run-ready: no (Hugging Bay has not certified a run path for this model; hosted bytes are downloadable and hash-checkable only)
- Peer-assisted fallbacks: 0 verified, 0 pending review
- Downloads: 276876
- Stars: 3
- Card completeness: 96%

## Description

This model is a quantized version of Llama 3.2 3B Instruct with FP8 weights and FP8 activations, and the card says it was obtained by quantizing the weights and activations to FP8. It is intended for commercial and research use in multiple languages, and it is suited to assistant like chat.

## Model Card Facts

- Declared license: llama3.2
- Pipeline: text-generation
- Library: unknown
- Base model: meta-llama/Llama-3.2-3B-Instruct
- Task categories: text-generation
- Languages: en, de, fr, it, pt, hi, es, th
- File hints: 9 upstream file samples

## File Hints

- .gitattributes (1.5 KB)
- README.md (8.6 KB)
- config.json (2.1 KB)
- generation_config.json (184 B)
- model.safetensors (4.1 GB) LFS
- recipe.yaml (351 B)
- special_tokens_map.json (296 B)
- tokenizer.json (16 MB) LFS
- tokenizer_config.json (53 KB)

## Quality Checklist

- Summary: passed
- Description: passed
- Model card metadata: passed
- License: review
- Provenance: passed
- File inventory: passed
- Hosted bytes: passed
- Scan result: passed
- Manifest signature: passed

## Safety Notes

- External catalog metadata imported from Hugging Face public API on 2026-09-20T22:04:49.143Z.
- Upstream license: llama3.2. Mirror admission policy: permissive (owner policy; no license rights inferred).
- Not a run kit: hosted, hash-checked bytes only; run readiness is not certified for this model yet.

## Actions

- Web: https://huggingbay.xyz/artifact/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic
- Files API: https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/files?limit=100
- Metadata: https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/metadata.json
- Distribution: https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/distribution
- Manifest: https://huggingbay.xyz/api/manifests/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic
- Download plan (per-file SHA-256 + signed download URLs): https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/download-plan
- Hosted download rows: https://huggingbay.xyz/api/downloadable-files?artifactId=hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic&limit=100
- Request mirror: not applicable.
- Submit peer fallback for review: POST https://huggingbay.xyz/api/artifacts/hf-model-redhatai-llama-3-2-3b-instruct-fp8-dynamic/decentralized-fallbacks