nemotron-3.5-super:120b-a12b-mlx-bf16

23 2 minutes ago

Nemotron 3.5 Super is NVIDIA’s vision-language model for coding, reasoning, tool use, and document understanding

vision tools thinking 120b
ce813d598192 · 1.1kB
{
"audio_hop_length": 160,
"audio_sampling_rate": 16000,
"audio_subsampling_factor": 8,
"auto_map": {
"AutoProcessor": "processing_nemotron_h_omni.NemotronH_Omni_Reasoning_V3Processor"
},
"image_processor": {
"_downsample_factor": 2,
"auto_map": {
"AutoImageProcessor": "image_processing_nemotron_h_omni.NemotronH_Omni_Reasoning_V3ImageProcessor",
"AutoProcessor": "processing_nemotron_h_omni.NemotronH_Omni_Reasoning_V3Processor"
},
"downsample_ratio": 0.5,
"image_processor_type": "NemotronH_Omni_Reasoning_V3ImageProcessor",
"max_model_len": 16384,
"max_num_patches": 13312,
"min_num_patches": 1024,
"norm_mean": [
0.48145466,
0.4578275,
0.40821073
],
"norm_std": [
0.26862954,
0.26130258,
0.27577711
],
"patch_size": 16,
"rescale_factor": 0.00392156862745098,
"video_maintain_aspect_ratio": true,
"video_target_num_patches": 1024
},
"processor_class": "NemotronH_Omni_Reasoning_V3Processor",
"video_temporal_patch_dim": 2
}