Hugging Face's logo

immanuelpeter
/
MiniMax-M3.1-Flash-Vision

Image Feature Extraction
Transformers
Safetensors
minimax_m3_vl_vision
vision
Model card Files
xet
 Community
immanuelpeter commited on
Commit
7a729a3
·
verified ·
1 Parent(s): 7371313

Drop the bench-evidence sentence from the card

Browse files
Files changed (2) hide show
  1. README.md +0 -1
  2. export_manifest.json +2 -2
README.md CHANGED
@@ -48,7 +48,6 @@ Forward one image per call: the native HF attention couples images in a single c
48
  The release tests compare all 515 Tower tensors and eight Projector tensors with the
49
  pinned parent checkpoint using `torch.equal`. Fixed-image Tower, merged and projected
50
  outputs also match the native parent implementation at 56 and 448 pixels.
51
- Full checks: [bench evidence](https://github.com/immanuel-peter/vision-tower-bench/tree/main/results/diagnostics/minimax-m31-flash).
52
 
53
  ## Reproduction
54
 
 
48
  The release tests compare all 515 Tower tensors and eight Projector tensors with the
49
  pinned parent checkpoint using `torch.equal`. Fixed-image Tower, merged and projected
50
  outputs also match the native parent implementation at 56 and 448 pixels.
 
51
 
52
  ## Reproduction
53
 
export_manifest.json CHANGED
@@ -17,7 +17,7 @@
17
  "LICENSE": "8c4978e3e0fe0349ff929cc8450463cba3039c4c9fc144d8bd5382ab00df247a",
18
  "LICENSE-CODE": "149f05ce186198084c9abd6af8e76e108ba7e507446a19ee5a2ffd694f69f525",
19
  "NOTICE": "19d19379d439b6dc2614476bb50b170fbe382a58f3a84ec67ff9bcd4e212d626",
20
- "README.md": "87f997d364c4c9f58bb309f59cd4a793d2cfecf92228d0d702f54541def440a7",
21
  "config.json": "9c3a6dac0654055594fdde1f148400331f3e55cb72ffe1b7a6806db07ab93a79",
22
  "examples/inference.py": "fcde5470ba4e9fce0730f544f4f3b7669bf4f9bdcaf556223cebc4ed67ef1c5a",
23
  "model.safetensors": "95e3731e26887d066a0b2291a1574da40a36aba5ba43e476746b9c16a64bcf08",
@@ -31,7 +31,7 @@
31
  "LICENSE": "8c4978e3e0fe0349ff929cc8450463cba3039c4c9fc144d8bd5382ab00df247a",
32
  "LICENSE-CODE": "149f05ce186198084c9abd6af8e76e108ba7e507446a19ee5a2ffd694f69f525",
33
  "NOTICE": "19d19379d439b6dc2614476bb50b170fbe382a58f3a84ec67ff9bcd4e212d626",
34
- "README.md": "87f997d364c4c9f58bb309f59cd4a793d2cfecf92228d0d702f54541def440a7",
35
  "config.json": "9c3a6dac0654055594fdde1f148400331f3e55cb72ffe1b7a6806db07ab93a79",
36
  "examples/inference.py": "fcde5470ba4e9fce0730f544f4f3b7669bf4f9bdcaf556223cebc4ed67ef1c5a",
37
  "model.safetensors": "95e3731e26887d066a0b2291a1574da40a36aba5ba43e476746b9c16a64bcf08",
 
17
  "LICENSE": "8c4978e3e0fe0349ff929cc8450463cba3039c4c9fc144d8bd5382ab00df247a",
18
  "LICENSE-CODE": "149f05ce186198084c9abd6af8e76e108ba7e507446a19ee5a2ffd694f69f525",
19
  "NOTICE": "19d19379d439b6dc2614476bb50b170fbe382a58f3a84ec67ff9bcd4e212d626",
20
+ "README.md": "4c85532cd87301b8ae8ec2686c6b3680aedda6d705089456ad7d65c701f3c6cc",
21
  "config.json": "9c3a6dac0654055594fdde1f148400331f3e55cb72ffe1b7a6806db07ab93a79",
22
  "examples/inference.py": "fcde5470ba4e9fce0730f544f4f3b7669bf4f9bdcaf556223cebc4ed67ef1c5a",
23
  "model.safetensors": "95e3731e26887d066a0b2291a1574da40a36aba5ba43e476746b9c16a64bcf08",
 
31
  "LICENSE": "8c4978e3e0fe0349ff929cc8450463cba3039c4c9fc144d8bd5382ab00df247a",
32
  "LICENSE-CODE": "149f05ce186198084c9abd6af8e76e108ba7e507446a19ee5a2ffd694f69f525",
33
  "NOTICE": "19d19379d439b6dc2614476bb50b170fbe382a58f3a84ec67ff9bcd4e212d626",
34
+ "README.md": "4c85532cd87301b8ae8ec2686c6b3680aedda6d705089456ad7d65c701f3c6cc",
35
  "config.json": "9c3a6dac0654055594fdde1f148400331f3e55cb72ffe1b7a6806db07ab93a79",
36
  "examples/inference.py": "fcde5470ba4e9fce0730f544f4f3b7669bf4f9bdcaf556223cebc4ed67ef1c5a",
37
  "model.safetensors": "95e3731e26887d066a0b2291a1574da40a36aba5ba43e476746b9c16a64bcf08",