Ministral-3-3B-Instruct-2512 (CI mirror)

This repository is an unmodified redistribution of mistralai/Ministral-3-3B-Instruct-2512 at upstream revision b35d4dfe56c142746f54dbd64f579faab2744308, published by Positron AI for internal CI and testing (a pinned, immutable source for automated model-ingest trace generation). It is not a Positron AI model, it carries no evaluation results, and it is not endorsed by Mistral AI. All credit for the model belongs to the upstream authors.

This repository exists solely as a pinned CI download source. It is not an acquisition channel for model processing: internal pipelines acquire this model from the canonical upstream repository (mistralai/Ministral-3-3B-Instruct-2512), and this mirror must never be referenced as a source or provenance identity in their place.

Per Apache License 2.0 section 4(b), the modification relative to upstream is: README.md (the upstream model card) was moved to README.upstream.md and replaced by this file. Every other file is byte-identical to the upstream revision above; the original model card, including its license declaration and usage notes, is preserved verbatim in README.upstream.md.

Provenance

Each payload file was verified against the upstream revision's Git LFS object IDs (sha256) and Git blob IDs before upload. sha256 of every file as published here:

file sha256
.gitattributes 1e11e8e6b9eb852a71eadd8fe191d4b9e329eb81308602dec19ab16da7e172b2
README.upstream.md 57623b9703d4fe40fe752b4a63f920c65ffd30cd7c281c8bccab7ee784460b67
SYSTEM_PROMPT.txt c6a46ef1b8017c5aef362480636b8c0046a9951acf338d3d8931bc1a3620315e
chat_template.jinja 0701cfbdc2b7d44fdbad104dff604faee4b0543e8247624568777fe465746f9b
config.json a9c3658d7bb45ab5be77e660d483a9d3f8a3d54bda2c69a1c576bdfefbc16be6
consolidated.safetensors a1b2aa6d22874ed04b7071a595581d48832ddeda8d6e69f54537b31aa3a775cf
generation_config.json e0923390059f84a9180b00e5501778acc45ea9856cd7f2fd68208b360927c677
model.safetensors 728f1826cd0e38191ca7b1379e81f78cf0555c6ffd95882aabd2404632346f86
params.json 184afbdca12f4075121650335d65ada27b471b8b2952c8f2b699c16516306459
processor_config.json ece2373c2ae391bce18785a1810543bc6173a1f28b6767cffab2c35dbea5f002
tekken.json 600bb27946565481ecf51ba8aee252e49b9a68507866080ac9c30185bb312843
tokenizer.json 99cf274236c60277fcfad861a5a1007518687ad06ba8938760f50b55ffa0b1ef
tokenizer_config.json 4a05af595aa11c3436d3b68a616219ae99b303426f708a12b31bc9ce32530728
Downloads last month
23
Safetensors
Model size
4B params
Tensor type
BF16
·
F8_E4M3
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for positron-ai/Ministral-3-3B-Instruct-2512