-
add video_processor.py
-
add tokenizer.json
-
add tokenizer_config.json
-
add quantization_config.json
-
add processing_minimax.py
-
add preprocessor_config.json
-
add model.safetensors.index.json
-
add model_extra_tensors.safetensors
-
add model-00035-of-00035.safetensors
-
add model-00034-of-00035.safetensors
-
add model-00033-of-00035.safetensors
-
add model-00032-of-00035.safetensors
-
add model-00031-of-00035.safetensors
-
add model-00030-of-00035.safetensors
-
add model-00029-of-00035.safetensors
-
add model-00028-of-00035.safetensors
-
add model-00027-of-00035.safetensors
-
Plain-language summary: drop rank/TP jargon + indexer wording (reader-friendly)
-
add model-00026-of-00035.safetensors
-
Throughput: drop internal jargon (eager mode) — plain number + hardware like Nex card
-
Clarify throughput in results table (~13 tok/s, eager) — match Nex card style
-
add model-00025-of-00035.safetensors
-
add model-00024-of-00035.safetensors
-
add model-00023-of-00035.safetensors
-
add model-00022-of-00035.safetensors
-
add model-00021-of-00035.safetensors
-
add model-00020-of-00035.safetensors
-
add model-00019-of-00035.safetensors
-
add model-00018-of-00035.safetensors
-
add model-00017-of-00035.safetensors
-
add model-00016-of-00035.safetensors
-
add model-00015-of-00035.safetensors
-
add model-00014-of-00035.safetensors
-
add model-00013-of-00035.safetensors
-
add model-00012-of-00035.safetensors
-
add model-00011-of-00035.safetensors
-
add model-00010-of-00035.safetensors
-
add model-00009-of-00035.safetensors
-
add model-00008-of-00035.safetensors
-
add model-00007-of-00035.safetensors
-
add model-00006-of-00035.safetensors
-
add model-00005-of-00035.safetensors
-
add model-00004-of-00035.safetensors
-
add model-00003-of-00035.safetensors
-
add model-00002-of-00035.safetensors
-
add model-00001-of-00035.safetensors
-
add image_processor.py
-
add generation_config.json
-
add configuration_minimax_m3_vl.py
-
add config.json
