Upload folder using huggingface_hub
Browse files- onnx_split/image_encoder.onnx +3 -0
- onnx_split/merges.txt +0 -0
- onnx_split/preprocess.json +17 -0
- onnx_split/special_tokens_map.json +30 -0
- onnx_split/text_encoder.onnx +3 -0
- onnx_split/tokenizer.json +0 -0
- onnx_split/tokenizer_config.json +33 -0
- onnx_split/vocab.json +0 -0
onnx_split/image_encoder.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbcdd55a8382093256cd9c6009f75d8018648edb15a2d16853fb33add387f250
|
| 3 |
+
size 143667032
|
onnx_split/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
onnx_split/preprocess.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model": "mobileclip2-s2",
|
| 3 |
+
"source": "open_clip MobileCLIP2-S2 pretrained=dfndr2b (timm/MobileCLIP2-S2-OpenCLIP)",
|
| 4 |
+
"image_size": 256,
|
| 5 |
+
"mean": [0.0, 0.0, 0.0],
|
| 6 |
+
"std": [1.0, 1.0, 1.0],
|
| 7 |
+
"interpolation": "bilinear",
|
| 8 |
+
"resize_mode": "shortest",
|
| 9 |
+
"context_len": 77,
|
| 10 |
+
"id_dtype": "int64",
|
| 11 |
+
"dim": 512,
|
| 12 |
+
"image_input": "pixel_values",
|
| 13 |
+
"image_output": "image_features",
|
| 14 |
+
"text_input": "input_ids",
|
| 15 |
+
"text_output": "text_features",
|
| 16 |
+
"note": "dfndr2b uses mean=0/std=1 (pixel/255 only), NOT OpenAI CLIP mean/std. Do not apply CLIP normalization."
|
| 17 |
+
}
|
onnx_split/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|startoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<|endoftext|>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<|endoftext|>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
onnx_split/text_encoder.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3503debcf8daf9cf6d5af069a9ec082de0355a62dce487f2501d6afd28b4e656
|
| 3 |
+
size 254083818
|
onnx_split/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
onnx_split/tokenizer_config.json
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"49406": {
|
| 5 |
+
"content": "<|startoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"49407": {
|
| 13 |
+
"content": "<|endoftext|>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
}
|
| 20 |
+
},
|
| 21 |
+
"backend": "tokenizers",
|
| 22 |
+
"bos_token": "<|startoftext|>",
|
| 23 |
+
"clean_up_tokenization_spaces": false,
|
| 24 |
+
"do_lower_case": true,
|
| 25 |
+
"eos_token": "<|endoftext|>",
|
| 26 |
+
"errors": "replace",
|
| 27 |
+
"extra_special_tokens": {},
|
| 28 |
+
"is_local": false,
|
| 29 |
+
"model_max_length": 77,
|
| 30 |
+
"pad_token": "<|endoftext|>",
|
| 31 |
+
"tokenizer_class": "CLIPTokenizer",
|
| 32 |
+
"unk_token": "<|endoftext|>"
|
| 33 |
+
}
|
onnx_split/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|