{
    "model": {
        "name": "MiniMax H3 FL2VA Pruned Veda Sparse 20B",
        "architecture": "minimax_h3_fl2va_pruned",
        "description": "Pruned First/Last-Frame-to-Video-and-Audio: generate synchronized video and stereo audio from text alone, a first frame, a last frame, or both. This variant uses Veda attention, which can be up to 2x faster when generating videos at higher resolutions or with longer durations.",
        "URLs": [
            "https://huggingface.co/DeepBeepMeep/MiniMax-H3/resolve/main/MiniMax-H3-FL2VA-pruned_rank8_bf16.safetensors",
            "https://huggingface.co/DeepBeepMeep/MiniMax-H3/resolve/main/MiniMax-H3-FL2VA-pruned_rank8_int8_convrot.safetensors"
        ],
        "veda": true,
        "modules": [
            [
                "https://huggingface.co/DeepBeepMeep/MiniMax-H3/resolve/main/minimax_h3/minimax_h3_t2va_veda_predictor_bf16.safetensors|minimax_h3",
                "https://huggingface.co/DeepBeepMeep/MiniMax-H3/resolve/main/minimax_h3/minimax_h3_t2va_veda_predictor_int8_convrot.safetensors|minimax_h3"
            ]
        ]
    },
    "overrides": "attention_mode='veda'",
    "prompt": "integrated_multimodal_description: [Shot 1] A five-second cinematic single take inside a rain-lashed glass observatory at midnight. A radio astronomer sits before an antique brass receiver while electric-blue constellations drift across a translucent star map behind her. The camera makes a slow push from medium shot to close-up as the receiver emits three precise tones. She leans toward its microphone, looks up with sudden hope, and says clearly (S1) <d>[English] If you can hear me, follow this signal.</d> While finishing the line, she presses the transmit key; one brilliant blue pulse races across the star map and lights her restrained smile. Her speech completes naturally within the shot with precise lip synchronization.\noverall_soundscape: Rain rattling across the observatory dome, a low electrical hum, three clean receiver tones, a brass switch clicking, the astronomer's synchronized voice, and a short stereo transmission pulse.\nnon_diegetic_music: One quiet bowed-glass chord rising beneath the transmission and fading at the end."
}