Instructions to use immanuelpeter/MiniMax-M3-Vision with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use immanuelpeter/MiniMax-M3-Vision with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-feature-extraction", model="immanuelpeter/MiniMax-M3-Vision")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("immanuelpeter/MiniMax-M3-Vision") model = AutoModel.from_pretrained("immanuelpeter/MiniMax-M3-Vision", device_map="auto") - Notebooks
- Google Colab
- Kaggle
File size: 893 Bytes
accba69 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 | {
"architectures": [
"MiniMaxM3VLVisionModel"
],
"attention_dropout": 0.0,
"dtype": "bfloat16",
"hidden_act": "gelu",
"hidden_size": 1280,
"image_size": 2016,
"img_token_compression_config": {
"image_token_compression_method": "patch_merge",
"spatial_merge_size": 2,
"temporal_patch_size": 2
},
"initializer_factor": 1.0,
"initializer_range": 0.02,
"intermediate_size": 5120,
"layer_norm_eps": 1e-05,
"model_type": "minimax_m3_vl_vision",
"num_attention_heads": 16,
"num_channels": 3,
"num_hidden_layers": 32,
"patch_size": 14,
"position_embedding_type": "rope",
"projection_dim": 6144,
"rope_mode": "3d",
"rope_parameters": {
"rope_theta": 10000.0,
"rope_type": "axial"
},
"spatial_merge_size": 2,
"temporal_patch_size": 2,
"transformers_version": "5.17.0",
"vision_segment_max_frames": 4,
"vocab_size": 32000
}
|