midigpt

A transformer model for computer-assisted multitrack music composition.

Fill in missing bars while preserving your existing arrangement
Generate new tracks from scratch, conditioned on musical attributes
Steer the output by controlling note density, polyphony, and note duration — globally or per bar
Run a MIDI-GPT API server via a HTTP server
One-line setup — load pretrained models from HuggingFace Hub, no compiler needed

Installation

pip install "midigpt[inference]"

Pre-built wheels for CPython 3.10–3.12 on Linux (x86_64), macOS (x86_64 + arm64), and Windows (AMD64). No compiler needed.

Extra	What it adds
`inference`	`torch>=2.0`, `tqdm`, `huggingface_hub`, `safetensors`
`train`	PyTorch Lightning, HuggingFace `datasets`, `pyarrow`, `python-dotenv`, `wandb`
`realtime`	`python-osc`, Flask, Flask-SocketIO
`http`	FastAPI, uvicorn
`dev`	`pytest`, `ruff`, `mypy`
`all`	`realtime` + `train`

Quick start

Load a pretrained model from HuggingFace Hub and generate music in four lines:

from midigpt import Score, Track, Bar
from midigpt.inference import InferenceEngine, GenerationRequest, InferenceConfig, TrackPrompt

engine = InferenceEngine.from_pretrained("yellow")

# 4-bar score with one empty melodic track
score = Score(tracks=[Track(bars=[Bar() for _ in range(4)])])

result = engine.session(
    score,
    GenerationRequest(
        tracks=[TrackPrompt(id=0, bars=[0, 1, 2, 3])],
        config=InferenceConfig(model_dim=4, mask_mode="attention"),
    ),
).run()

total = sum(len(b.notes) for t in result.tracks for b in t.bars)
print(f"Generated {total} notes")
result.to_midi("output.mid")

The model is downloaded once and cached by huggingface_hub in ~/.cache/huggingface/hub/.

Models

Name	`num_bars_map`	Infill	Attributes	Download
`yellow`	4, 8	yes	note density, polyphony (min/max), note duration (min/max)	yellow_medium-final.safetensors
`prism_medium`	4, 8, 12, 16	yes	key signature, pitch range, silence, note duration, note density (bar), polyphony (bar), pitch class set, genre	prism_medium-step58000.safetensors — training in progress
`expressive`	4, 8, 12, 16	yes	key signature, pitch range, silence, note duration, note density (bar), polyphony (bar), pitch class set, nomml, genre	expressive_medium-step56000.safetensors — training in progress
`ghost`	4, 8, 12, 16	yes	note density, polyphony (min/max), note duration (min/max)	not released — planned architecture, see docs/models.md#ghost

model_dim in InferenceConfig is the context window in bars, not a vocabulary dimension — pass a value from the model's num_bars_map. expressive additionally encodes sub-grid timing via delta tokens and supports switchable velocity/microtiming controls. See docs/models.md for the full breakdown.

Inference API

Load a model

# By name (downloads from Metacreation/MIDI-GPT on HuggingFace Hub)
engine = InferenceEngine.from_pretrained("yellow")   # or "prism_medium", "expressive"

# From a local checkpoint (.safetensors or .pt bundle)
engine = InferenceEngine.from_checkpoint("path/to/model.safetensors")

Infill existing bars

score = Score.from_midi("my_song.mid")

request = GenerationRequest(
    tracks=[
        TrackPrompt(id=0, bars=[4, 5, 6, 7]),   # bars to regenerate
        TrackPrompt(id=1, bars=[], ignore=True), # leave track 1 unchanged
    ],
    config=InferenceConfig(temperature=1.0, top_p=0.95, model_dim=8),
)

result = engine.session(score, request).run()
result.to_midi("output.mid")

Autoregressive generation from scratch

request = GenerationRequest(
    tracks=[
        TrackPrompt(
            id=0,
            bars=[],
            autoregressive=True,
            attributes={"max_polyphony": 3},      # quantized attribute level
            controls={"time_signature": 0},        # index into encoder TS list
        ),
    ],
    config=InferenceConfig(temperature=1.0, model_dim=8, polyphony_hard_limit=4),
)
result = engine.session(score, request).run()

Key types

Class	Module	Purpose
`InferenceEngine`	`midigpt.inference`	Top-level loader and session factory
`GenerationRequest`	`midigpt.inference`	Bundle of per-track prompts and config
`TrackPrompt`	`midigpt.inference`	Per-track bars, mode, attributes, controls
`InferenceConfig`	`midigpt.inference`	Temperature, sampling filters, step planner
`SamplingSession`	`midigpt.inference`	Token-level sampling loop (returned by `session()`)

`TrackPrompt` fields

Field	Type	Default	Meaning
`id`	int	—	Track index in the score
`bars`	list[int]	—	Bars to generate
`autoregressive`	bool	`False`	Generate from scratch (no per-bar prompt)
`ignore`	bool	`False`	Omit this track from the token stream
`mask_bars`	list[int]	`[]`	Bars hidden with `MASK_BAR` (disjoint from `bars`)
`attributes`	dict[str,int]	`{}`	Quantized attribute overrides
`controls`	dict[str,Any]	`{}`	Token locks e.g. `{"time_signature": 0}`
`bar_attributes`	dict[int,dict]	`{}`	Per-bar attribute overrides (absolute bar index)
`bar_controls`	dict[int,dict]	`{}`	Per-bar control overrides (absolute bar index)

Sampling filters

InferenceConfig exposes a four-stage logit-filtering pipeline (top_k → top_p → mask_k → mask_p):

Field	Default	Meaning
`top_k`	`0` (off)	Keep top-k highest-probability tokens
`top_p`	`1.0` (off)	Nucleus: keep the smallest set summing to ≥ `top_p`
`mask_k`	`0` (off)	Remove the top-k most-likely tokens (novelty pressure)
`mask_p`	`0.0` (off)	Anti-nucleus: remove tokens summing to ≥ `mask_p` from the top

A small mask_k=1 or mask_p=0.3 pushes the model off its most-confident picks — useful for getting diverse outputs when novelty_check=True.

Mask modes

Control how future bars appear in the context window:

Mode	Behaviour
`"token"`	Encoder emits a `MaskBar` token (requires vocab support)
`"attention"`	Future bars zeroed in the KV cache via exact span masking
`"attention_approx"`	Single prefill mask + KV surgery; cheaper than `"attention"`
`"attention_skip"`	Future tokens filtered from input; `position_ids` passed explicitly
`"remove"`	Future bars omitted entirely from the token stream

Set via InferenceConfig(mask_mode="attention"). "attention" works on all encoders; "token" requires the encoder vocab to include a MaskBar domain.

Attribute controls

Introspect available controls at runtime:

engine._analyzer.attribute_sizes()         # {"note_density": 10, "min_polyphony": 10, ...}
engine._analyzer.attribute_value_labels()  # {"note_density": ["very sparse", ...], ...}
engine._analyzer.attribute_track_types()   # {"note_density": "melodic", ...}

Pass quantized levels (integers in [0, size)) in TrackPrompt.attributes.

Training

1. Preprocess parquet shards

python -m midigpt.training.preprocess \
    --parquet /data/train/*.parquet \
    --checkpoint models/yellow_medium-final.safetensors

Builds a valid-index cache so dataset initialization is instant on subsequent runs. Cached in ~/.midigpt/ (override with MIDIGPT_CACHE).

2. Launch training

python -m midigpt.training.trainer \
    --config     models/train_config.json \
    --train-data /data/train/*.parquet \
    --eval-data  /data/valid/*.parquet \
    --output-dir checkpoints/run_001

3. Python API

from midigpt.training.trainer import TrainConfig, train

config = TrainConfig.from_file("models/train_config.json")
train(config, train_path="/data/train/00000.parquet", eval_path="/data/valid/00000.parquet")

train() uses PyTorch Lightning and writes a packed .safetensors bundle at the end of training containing weights, architecture config, and encoder config.

Key `TrainConfig` fields

Field	Default	Notes
`n_embd` / `n_layer` / `n_head`	`512 / 6 / 8`	Model architecture
`max_seq_len`	`2048`	Token sequence cap
`infill_probability`	`0.75`	Fraction of samples trained with FillIn tokens
`mask_apply_probability`	`0.5`	Fraction of samples with `MASK_BAR` applied
`precision`	`"fp16"`	`"fp16"`, `"bf16"`, or `"fp32"`
`logger`	`"none"`	`"tensorboard"`, `"wandb"`, or `"none"`
`num_workers`	`0`	Must be 0 — the C++ MIDI parser is not fork-safe

HTTP server

pip install "midigpt[http]"

# From a local checkpoint (.safetensors or .pt)
midigpt-http --ckpt checkpoints/run_001/model_final.safetensors --port 8000

# From HuggingFace Hub (by name or repo ID)
midigpt-http --pretrained yellow --port 8000
midigpt-http --pretrained Metacreation/MIDI-GPT --hf-filename yellow_medium-final.safetensors --port 8000

A stateless REST API — every request carries the full score and generation parameters. The interactive API docs are available at http://localhost:8000/docs.

Endpoint	Description
`GET /health`	Liveness probe
`GET /info`	Model capabilities and attribute sizes
`POST /generate`	`{score, request}` → `{score, timing}`

# Score: 1 melodic track, 4 empty bars — generate all 4 from scratch
curl -s -X POST http://localhost:8000/generate \
  -H "Content-Type: application/json" \
  -d '{
    "score": {
      "resolution": 480, "tempo": 500000,
      "tracks": [{
        "instrument": 0, "track_type": "melodic",
        "bars": [
          {"ts_numerator": 4, "ts_denominator": 4, "notes": []},
          {"ts_numerator": 4, "ts_denominator": 4, "notes": []},
          {"ts_numerator": 4, "ts_denominator": 4, "notes": []},
          {"ts_numerator": 4, "ts_denominator": 4, "notes": []}
        ]
      }]
    },
    "request": {
      "tracks": [{"id": 0, "bars": [0, 1, 2, 3]}],
      "config": {"model_dim": 4}
    }
  }' | jq .score

Use --device cuda, --device mps, or --device auto (default) to select the compute device.

Real-time OSC server

pip install "midigpt[realtime]"
midigpt-server --ckpt models/yellow_medium-final.safetensors --port 7400

Listens for OSC messages on a UDP port and streams generated notes back in real time. Generation is triggered bar-by-bar via /midigpt/bar/end on a background thread.

Selected OSC addresses:

Address	Direction	Description
`/midigpt/session/init`	in	Start a new session
`/midigpt/track/create`	in	Register a track
`/midigpt/note`	in	Push an incoming note
`/midigpt/bar/end`	in	Signal bar end (triggers generation)
`/midigpt/param/set`	in	Adjust sampling parameters at runtime
`/midigpt/attr/set`	in	Set attribute overrides
`/midigpt/generated/note`	out	Emit a generated note
`/midigpt/generated/features`	out	Per-bar statistics
`/midigpt/capabilities`	out	Attribute support for the loaded checkpoint

Development

Setup

git clone https://github.com/Metacreation-Lab/MIDI-GPT.git
cd MIDI-GPT
pip install -e ".[inference,dev]"   # compiles the C++ extension in-place

Prerequisites: Python 3.10+, CMake 3.21+, a C++20 compiler.

Tests

# Python
pytest tests/python/
pytest tests/python -m "not slow and not inference"   # CI subset (no model needed)

# C++
cmake -S . -B build_cpp -DCMAKE_BUILD_TYPE=Release
cmake --build build_cpp -j
ctest --test-dir build_cpp --output-on-failure

Linting

ruff check src/ tests/    # lint
ruff format src/ tests/   # format

pre-commit runs both automatically on commit:

pip install pre-commit && pre-commit install

Release

Tag a commit vX.Y.Z → .github/workflows/wheels.yml builds wheels on Linux / macOS / Windows × Python 3.10–3.12, drafts a GitHub Release, and publishes to PyPI via OIDC Trusted Publishing.

Logging

Set MIDIGPT_LOG_LEVEL=DEBUG (or a numeric level) before importing. Accepts both string names (DEBUG, INFO, WARNING) and integers.

Citation

@misc{pasquier2025midigptcontrollablegenerativemodel,
      title={MIDI-GPT: A Controllable Generative Model for Computer-Assisted Multitrack Music Composition},
      author={Philippe Pasquier and Jeff Ens and Nathan Fradet and Paul Triana and Davide Rizzotti and Jean-Baptiste Rolland and Maryam Safi},
      year={2025},
      eprint={2501.17011},
      archivePrefix={arXiv},
      primaryClass={cs.SD},
      url={https://arxiv.org/abs/2501.17011},
}

License

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

midigpt-0.3.2.tar.gz (9.0 MB view details)

Uploaded Jul 9, 2026 Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

midigpt-0.3.2-cp312-cp312-win_amd64.whl (1.1 MB view details)

Uploaded Jul 9, 2026 CPython 3.12Windows x86-64

midigpt-0.3.2-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (812.9 kB view details)

Uploaded Jul 9, 2026 CPython 3.12manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

midigpt-0.3.2-cp312-cp312-macosx_11_0_arm64.whl (591.1 kB view details)

Uploaded Jul 9, 2026 CPython 3.12macOS 11.0+ ARM64

midigpt-0.3.2-cp312-cp312-macosx_10_15_x86_64.whl (678.7 kB view details)

Uploaded Jul 9, 2026 CPython 3.12macOS 10.15+ x86-64

midigpt-0.3.2-cp311-cp311-win_amd64.whl (784.5 kB view details)

Uploaded Jul 9, 2026 CPython 3.11Windows x86-64

midigpt-0.3.2-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (811.8 kB view details)

Uploaded Jul 9, 2026 CPython 3.11manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

midigpt-0.3.2-cp311-cp311-macosx_11_0_arm64.whl (589.8 kB view details)

Uploaded Jul 9, 2026 CPython 3.11macOS 11.0+ ARM64

midigpt-0.3.2-cp311-cp311-macosx_10_15_x86_64.whl (674.8 kB view details)

Uploaded Jul 9, 2026 CPython 3.11macOS 10.15+ x86-64

midigpt-0.3.2-cp310-cp310-win_amd64.whl (459.7 kB view details)

Uploaded Jul 9, 2026 CPython 3.10Windows x86-64

midigpt-0.3.2-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (811.2 kB view details)

Uploaded Jul 9, 2026 CPython 3.10manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

midigpt-0.3.2-cp310-cp310-macosx_11_0_arm64.whl (587.8 kB view details)

Uploaded Jul 9, 2026 CPython 3.10macOS 11.0+ ARM64

midigpt-0.3.2-cp310-cp310-macosx_10_15_x86_64.whl (673.3 kB view details)

Uploaded Jul 9, 2026 CPython 3.10macOS 10.15+ x86-64

File details

Details for the file midigpt-0.3.2.tar.gz.

File metadata

Download URL: midigpt-0.3.2.tar.gz
Upload date: Jul 9, 2026
Size: 9.0 MB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2.tar.gz
Algorithm	Hash digest
SHA256	`6f309631842362e1ba693b3122eabd08bef7781bfe447e179b21274488ba74b9`
MD5	`355fccf5907f09c72fb000c817709fea`
BLAKE2b-256	`027722933d3006962ec8a053bb14ef3353c828dcf49ddd5a21185336e734d035`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2.tar.gz:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2.tar.gz
- Subject digest: 6f309631842362e1ba693b3122eabd08bef7781bfe447e179b21274488ba74b9
- Sigstore transparency entry: 2128019938
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp312-cp312-win_amd64.whl.

File metadata

Download URL: midigpt-0.3.2-cp312-cp312-win_amd64.whl
Upload date: Jul 9, 2026
Size: 1.1 MB
Tags: CPython 3.12, Windows x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp312-cp312-win_amd64.whl
Algorithm	Hash digest
SHA256	`019a4cf9200cd71fccaa38dfb28748c8c959f462c32961c86496615180a868ba`
MD5	`baa750d4cba3b2ed0da7a64e2daf2ab5`
BLAKE2b-256	`1cdfa8abf8df5cfaceaa0e26cd8afe2402e8c1d2b0dde01475db6ec89c4e8608`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp312-cp312-win_amd64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp312-cp312-win_amd64.whl
- Subject digest: 019a4cf9200cd71fccaa38dfb28748c8c959f462c32961c86496615180a868ba
- Sigstore transparency entry: 2128020099
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

Download URL: midigpt-0.3.2-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Upload date: Jul 9, 2026
Size: 812.9 kB
Tags: CPython 3.12, manylinux: glibc 2.27+ x86-64, manylinux: glibc 2.28+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm	Hash digest
SHA256	`871fc52376de4426cfb55436ae1fe617527ab8e0758603585c3a614af8960f7e`
MD5	`0744716c999a05e709f117845acdea19`
BLAKE2b-256	`b2a657747a335fda8f08cedd9d02c48ea273b6c512e5d03c37de0eda4c12d19d`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
- Subject digest: 871fc52376de4426cfb55436ae1fe617527ab8e0758603585c3a614af8960f7e
- Sigstore transparency entry: 2128020373
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp312-cp312-macosx_11_0_arm64.whl.

File metadata

Download URL: midigpt-0.3.2-cp312-cp312-macosx_11_0_arm64.whl
Upload date: Jul 9, 2026
Size: 591.1 kB
Tags: CPython 3.12, macOS 11.0+ ARM64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp312-cp312-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`c72afb576d7ce397017adde899a7cfac0aa9377c2d22ea9004270047d4058de9`
MD5	`fff76909766015db036a4484ddd72c15`
BLAKE2b-256	`e606d6e4a4f2cb685fa90553944822ead130c1039e55eb273b921e3c5424ab8e`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp312-cp312-macosx_11_0_arm64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp312-cp312-macosx_11_0_arm64.whl
- Subject digest: c72afb576d7ce397017adde899a7cfac0aa9377c2d22ea9004270047d4058de9
- Sigstore transparency entry: 2128020731
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp312-cp312-macosx_10_15_x86_64.whl.

File metadata

Download URL: midigpt-0.3.2-cp312-cp312-macosx_10_15_x86_64.whl
Upload date: Jul 9, 2026
Size: 678.7 kB
Tags: CPython 3.12, macOS 10.15+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp312-cp312-macosx_10_15_x86_64.whl
Algorithm	Hash digest
SHA256	`abbe8e2260e31a0e10a54b80591ca43fcbed81ab40979cf05f074de09626b4a6`
MD5	`fa63ea8e76ab73a960585575dce0cd32`
BLAKE2b-256	`657879d9bb49732926ff7a6cba4cdf7b3951802ede565ee3bd8602a5303ce29d`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp312-cp312-macosx_10_15_x86_64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp312-cp312-macosx_10_15_x86_64.whl
- Subject digest: abbe8e2260e31a0e10a54b80591ca43fcbed81ab40979cf05f074de09626b4a6
- Sigstore transparency entry: 2128021084
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp311-cp311-win_amd64.whl.

File metadata

Download URL: midigpt-0.3.2-cp311-cp311-win_amd64.whl
Upload date: Jul 9, 2026
Size: 784.5 kB
Tags: CPython 3.11, Windows x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp311-cp311-win_amd64.whl
Algorithm	Hash digest
SHA256	`8e46296eabe1812e72670630301ed77ae1fb672c58e532be3d178fc24bc1db9f`
MD5	`209b398da2192bdd76cf7450591bb7f4`
BLAKE2b-256	`79d36ee4d0bd168c2d9962f1490f38154f0dec4050aae521de1334c8dc1b2aa4`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp311-cp311-win_amd64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp311-cp311-win_amd64.whl
- Subject digest: 8e46296eabe1812e72670630301ed77ae1fb672c58e532be3d178fc24bc1db9f
- Sigstore transparency entry: 2128021338
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

Download URL: midigpt-0.3.2-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Upload date: Jul 9, 2026
Size: 811.8 kB
Tags: CPython 3.11, manylinux: glibc 2.27+ x86-64, manylinux: glibc 2.28+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm	Hash digest
SHA256	`a85e315ee4c8084e0efcf07723e6d50e7e4425f6a44072400d5222de0bdfe567`
MD5	`172d9afa4ccda9d3bef78c87ad7cecf6`
BLAKE2b-256	`b8789eae5e0876709b3117c1a1cc77969477a1e6b44df79b89b538232db906e8`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
- Subject digest: a85e315ee4c8084e0efcf07723e6d50e7e4425f6a44072400d5222de0bdfe567
- Sigstore transparency entry: 2128020626
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp311-cp311-macosx_11_0_arm64.whl.

File metadata

Download URL: midigpt-0.3.2-cp311-cp311-macosx_11_0_arm64.whl
Upload date: Jul 9, 2026
Size: 589.8 kB
Tags: CPython 3.11, macOS 11.0+ ARM64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp311-cp311-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`062a26db212b164ab228e1ef2220ac79bfe873ad0b436272f04c1841e1830067`
MD5	`314fbe406ba78f2b4fb646285f8e7f1e`
BLAKE2b-256	`a4f136ecf8b0841a3d0040a8f1085e1c87e52982578a89752bf4c531c4edc7ed`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp311-cp311-macosx_11_0_arm64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp311-cp311-macosx_11_0_arm64.whl
- Subject digest: 062a26db212b164ab228e1ef2220ac79bfe873ad0b436272f04c1841e1830067
- Sigstore transparency entry: 2128020836
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp311-cp311-macosx_10_15_x86_64.whl.

File metadata

Download URL: midigpt-0.3.2-cp311-cp311-macosx_10_15_x86_64.whl
Upload date: Jul 9, 2026
Size: 674.8 kB
Tags: CPython 3.11, macOS 10.15+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp311-cp311-macosx_10_15_x86_64.whl
Algorithm	Hash digest
SHA256	`76d728a74a7871b3557232f187760768a2430c506ad53961c4984bd31aa86852`
MD5	`01a2ba47f2154c96f35e9770dd9585c8`
BLAKE2b-256	`60481cc3e942e76ed51a2d33950324025b02a1e5ea7e0bf4f69c42bd39bdda5c`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp311-cp311-macosx_10_15_x86_64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp311-cp311-macosx_10_15_x86_64.whl
- Subject digest: 76d728a74a7871b3557232f187760768a2430c506ad53961c4984bd31aa86852
- Sigstore transparency entry: 2128021431
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp310-cp310-win_amd64.whl.

File metadata

Download URL: midigpt-0.3.2-cp310-cp310-win_amd64.whl
Upload date: Jul 9, 2026
Size: 459.7 kB
Tags: CPython 3.10, Windows x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp310-cp310-win_amd64.whl
Algorithm	Hash digest
SHA256	`e7822a41e3281dcfbaa8192700f200e893a704ed32752972f09d4266a43f40d0`
MD5	`2fd5aa25224d8d8cdaf26d8f80fa4bd9`
BLAKE2b-256	`f95eabbf7e39056955831fe3d0d2dad2b19fa3851ed5836e36ff100a2cca8350`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp310-cp310-win_amd64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp310-cp310-win_amd64.whl
- Subject digest: e7822a41e3281dcfbaa8192700f200e893a704ed32752972f09d4266a43f40d0
- Sigstore transparency entry: 2128020516
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

Download URL: midigpt-0.3.2-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Upload date: Jul 9, 2026
Size: 811.2 kB
Tags: CPython 3.10, manylinux: glibc 2.27+ x86-64, manylinux: glibc 2.28+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm	Hash digest
SHA256	`045c76aeb5c9c5876d79af66a48d2f9e946d371048c2812799c6944a5c11fe75`
MD5	`db51a31b32564601b088753b32763579`
BLAKE2b-256	`f945604eff5f8f58354ceb7a1992336ded11e251408a0647da67a08645759c79`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
- Subject digest: 045c76aeb5c9c5876d79af66a48d2f9e946d371048c2812799c6944a5c11fe75
- Sigstore transparency entry: 2128020262
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp310-cp310-macosx_11_0_arm64.whl.

File metadata

Download URL: midigpt-0.3.2-cp310-cp310-macosx_11_0_arm64.whl
Upload date: Jul 9, 2026
Size: 587.8 kB
Tags: CPython 3.10, macOS 11.0+ ARM64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp310-cp310-macosx_11_0_arm64.whl
Algorithm	Hash digest
SHA256	`b2cfdca7b4f01491326c9effff11311dc3136c576d467fb212f8972fa25d7d11`
MD5	`875e97b954c52975cc5ddaad98818728`
BLAKE2b-256	`99e03d43027d3a17d964c5c7756a2c37f7678c4a1c2630d11ee6499606482158`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp310-cp310-macosx_11_0_arm64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp310-cp310-macosx_11_0_arm64.whl
- Subject digest: b2cfdca7b4f01491326c9effff11311dc3136c576d467fb212f8972fa25d7d11
- Sigstore transparency entry: 2128021224
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

File details

Details for the file midigpt-0.3.2-cp310-cp310-macosx_10_15_x86_64.whl.

File metadata

Download URL: midigpt-0.3.2-cp310-cp310-macosx_10_15_x86_64.whl
Upload date: Jul 9, 2026
Size: 673.3 kB
Tags: CPython 3.10, macOS 10.15+ x86-64
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for midigpt-0.3.2-cp310-cp310-macosx_10_15_x86_64.whl
Algorithm	Hash digest
SHA256	`851e33db4b48dc90173f77f3129d022d2793c3d53da197aa675920adee657834`
MD5	`bd451ab5e46133cbd2bcffcfb3701267`
BLAKE2b-256	`61fe8fc2354796f904bfe6a1fa6c5ccccacf3127685eb5f185629b6dda765923`

See more details on using hashes here.

Provenance

The following attestation bundles were made for midigpt-0.3.2-cp310-cp310-macosx_10_15_x86_64.whl:

Publisher: wheels.yml on Metacreation-Lab/MIDI-GPT

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: midigpt-0.3.2-cp310-cp310-macosx_10_15_x86_64.whl
- Subject digest: 851e33db4b48dc90173f77f3129d022d2793c3d53da197aa675920adee657834
- Sigstore transparency entry: 2128020940
- Sigstore integration time: Jul 9, 2026
Source repository:
- Permalink: Metacreation-Lab/MIDI-GPT@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Branch / Tag: refs/tags/v0.3.2
- Owner: https://github.com/Metacreation-Lab
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: wheels.yml@2b5310eb3febb25ed833a3b5da8a868f944f4540
- Trigger Event: push

Release history Release notifications | RSS feed

0.3.3

Jul 30, 2026

This release

0.3.2

Jul 9, 2026

0.3.1

Jul 6, 2026

0.3.0

Jul 6, 2026

0.2.4

Jun 3, 2026

0.2.3

Jun 1, 2026

0.2.2 yanked

Jun 1, 2026

0.2.1 yanked

May 19, 2026

midigpt 0.3.2

midigpt

Installation

Quick start

Models

Inference API

Load a model

Infill existing bars

Autoregressive generation from scratch

Key types

TrackPrompt fields

Sampling filters

Mask modes

Attribute controls

Training

1. Preprocess parquet shards

2. Launch training

3. Python API

Key TrainConfig fields

HTTP server

Real-time OSC server

Development

Setup

Tests

Linting

Release

Logging

Citation

License

Project links

Key dates

1 maintainer

Credits

GitHub Statistics

License

Requires

Provides Extra

Classifiers

Download files

Source Distribution

Built Distributions

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

`TrackPrompt` fields

Key `TrainConfig` fields