AI Potluck
Contribute
Back to Gap Map Model components / Fine-tuned / chat models

Granite Code Instruct

IBM
open weights / Overall score: 1.7

IBM's enterprise code model line, fine-tuned from Granite Code Base for code instruction-following at 3B, 8B, 20B and 34B with 2K and 128K context variants. It was trained on permissively licensed code, math and natural-language instruction data, which is IBM's stated basis for offering enterprises clarity on training-data provenance.

The model card opens with a deprecation warning that advises against use in new projects.

Openness

3 high confidence
3.0
weights
open(Apache-2.0,on HF)
code
partial(inference + fine-tune sample scripts via Dolomite Engine
data
documented-not-released(CommitPackFT/MathInstruct/Glaive/HelpSteer/Open-Platypus named in card but processed corpus not redistributable)
post-training-data
described(SFT mixture named, not released)
paper
open(arXiv 2405.04324)
model_card
open
license
Apache-2.0(OSI)

Granite Code Instruct carries an OSI Apache-2.0 license with downloadable weights, a paper and a model card, but the training and post-training data are only described, not released, and no full training pipeline ships. That keeps it at open weights rather than open source, though it is genuinely permissive with no use caps.

  • https://github.com/ibm-granite/granite-code-models recorded 2026-08-13

    repo with Apache-2.0, fine-tune sample scripts (Dolomite Engine), but no training data download and no full training pipeline

  • https://huggingface.co/ibm-granite/granite-3b-code-instruct-2k recorded 2026-08-13

    `license:apache-2.0` in the repo tags and `"gated":false`; training sets NAMED as dataset tags (bigcode/commitpackft, TIGER-Lab/MathInstruct, meta-math/MetaMathQA, glaiveai/glaive-code-assistant-v3, garage-bAInd/Open-Platypus, nvidia/HelpSteer) with no processed mixture published; arxiv:2405.04324; model-index with HumanEvalSynthesis scores; a DEPRECATION WARNING banner at the top of the card

  • https://research.ibm.com/blog/granite-code-models-open-source recorded 2026-08-13

    IBM describes data sources (GitHub Code Clean, StarCoderData, CodeNet) but does not release the processed datasets

Adoption

1 medium confidence
1.0

IBM has deprecated Granite Code Instruct in favor of Granite 4.x, and download volume across the family's Instruct sizes has fallen to hobbyist and legacy-research scale. Both declared checkpoints show minimal recent pull activity, consistent with an end-of-life model rather than one in active production use.

Capability

2 high confidence
2.0

The 3B instruct model scores about 51% pass@1 on HumanEval Python, far below current frontier coding models.

Verified 2026-09-27, except some scores, still awaiting re-confirmation