Text-to-Image
MLX
MiniMax H3
Safetensors
English
apple-silicon
macos
image-to-image
image-editing
image-generation
Instructions to use devin-lai/MiniMax-H3-Image-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use devin-lai/MiniMax-H3-Image-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download devin-lai/MiniMax-H3-Image-mlx --local-dir MiniMax-H3-Image-mlx
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download optimization/round-03-compact/contract.json from devin-lai/MiniMax-H3-Image-mlx: direct link, hf CLI and curl.
- Browser
- Download file 1.26 kB
-
https://huggingface.co/devin-lai/MiniMax-H3-Image-mlx/resolve/main/optimization/round-03-compact/contract.json
- Command line
-
hf download hf://devin-lai/MiniMax-H3-Image-mlx/optimization/round-03-compact/contract.json
-
curl -L -o contract.json https://huggingface.co/devin-lai/MiniMax-H3-Image-mlx/resolve/main/optimization/round-03-compact/contract.json
1.26 kB
| { | |
| "schema_version": 1, | |
| "campaign_id": "h3-m5pro-round-03-compact", | |
| "objective": "latency", | |
| "quality_cases": [ | |
| "product", | |
| "portrait", | |
| "edit", | |
| "holdout-detail" | |
| ], | |
| "timing_cases": [ | |
| "product", | |
| "portrait" | |
| ], | |
| "quality_gates": { | |
| "max_abs_error": { | |
| "direction": "max", | |
| "limit": 0 | |
| }, | |
| "nonfinite_values": { | |
| "direction": "max", | |
| "limit": 0 | |
| } | |
| }, | |
| "workload_manifest_sha256": "8d14bb9873982547f69bbbd13e2a1e645f97454f14428220aa78001d4f843b13", | |
| "reference_sha256": "5fccbc8664a6efb3f9ff7e14b764c1aae876785fc28fd492dbb6c24b1b773a8b", | |
| "evaluator_sha256": "4db0c5e8606ffc011bf85dbadabc482468a2e49cd5a1757a774620d493eba7ee", | |
| "min_repeats": 3, | |
| "min_latency_reduction_pct": 2, | |
| "min_memory_reduction_pct": 3, | |
| "max_latency_regression_pct": 0, | |
| "max_memory_regression_pct": 5, | |
| "peak_memory_limit_bytes": 42949672960, | |
| "search_budget": "At most 40 minutes and two complete configurations: current packed12 champion vs same storage plus exact fusions/FC2. Same 4 quality and 2 timing cases, three paired warm repeats, <=1.5 GiB additional disk. One GPU job at a time. Start after dense campaign, global experiment deadline 10:02 UTC. Reused validation cases, not fresh holdouts." | |
| } | |