Commit
•
c2de364
1
Parent(s):
0524afe
Upload processor
Browse files- preprocessor_config.json +1 -0
- processor_config.json +6 -0
- tokenizer.json +8 -24
preprocessor_config.json
CHANGED
@@ -7,6 +7,7 @@
|
|
7 |
"do_center_crop": true,
|
8 |
"do_convert_rgb": true,
|
9 |
"do_normalize": true,
|
|
|
10 |
"do_rescale": true,
|
11 |
"do_resize": true,
|
12 |
"image_grid_pinpoints": [
|
|
|
7 |
"do_center_crop": true,
|
8 |
"do_convert_rgb": true,
|
9 |
"do_normalize": true,
|
10 |
+
"do_pad": true,
|
11 |
"do_rescale": true,
|
12 |
"do_resize": true,
|
13 |
"image_grid_pinpoints": [
|
processor_config.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"image_token": "<image>",
|
3 |
+
"patch_size": 14,
|
4 |
+
"processor_class": "LlavaNextProcessor",
|
5 |
+
"vision_feature_select_strategy": "default"
|
6 |
+
}
|
tokenizer.json
CHANGED
@@ -1,14 +1,7 @@
|
|
1 |
{
|
2 |
"version": "1.0",
|
3 |
"truncation": null,
|
4 |
-
"padding":
|
5 |
-
"strategy": "BatchLongest",
|
6 |
-
"direction": "Right",
|
7 |
-
"pad_to_multiple_of": null,
|
8 |
-
"pad_id": 0,
|
9 |
-
"pad_type_id": 0,
|
10 |
-
"pad_token": "<unk>"
|
11 |
-
},
|
12 |
"added_tokens": [
|
13 |
{
|
14 |
"id": 0,
|
@@ -47,23 +40,13 @@
|
|
47 |
"special": true
|
48 |
}
|
49 |
],
|
50 |
-
"normalizer":
|
51 |
-
|
52 |
-
"
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
},
|
57 |
-
{
|
58 |
-
"type": "Replace",
|
59 |
-
"pattern": {
|
60 |
-
"String": " "
|
61 |
-
},
|
62 |
-
"content": "▁"
|
63 |
-
}
|
64 |
-
]
|
65 |
},
|
66 |
-
"pre_tokenizer": null,
|
67 |
"post_processor": {
|
68 |
"type": "TemplateProcessing",
|
69 |
"single": [
|
@@ -150,6 +133,7 @@
|
|
150 |
"end_of_word_suffix": null,
|
151 |
"fuse_unk": true,
|
152 |
"byte_fallback": true,
|
|
|
153 |
"vocab": {
|
154 |
"<unk>": 0,
|
155 |
"<s>": 1,
|
|
|
1 |
{
|
2 |
"version": "1.0",
|
3 |
"truncation": null,
|
4 |
+
"padding": null,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
5 |
"added_tokens": [
|
6 |
{
|
7 |
"id": 0,
|
|
|
40 |
"special": true
|
41 |
}
|
42 |
],
|
43 |
+
"normalizer": null,
|
44 |
+
"pre_tokenizer": {
|
45 |
+
"type": "Metaspace",
|
46 |
+
"replacement": "▁",
|
47 |
+
"prepend_scheme": "first",
|
48 |
+
"split": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
49 |
},
|
|
|
50 |
"post_processor": {
|
51 |
"type": "TemplateProcessing",
|
52 |
"single": [
|
|
|
133 |
"end_of_word_suffix": null,
|
134 |
"fuse_unk": true,
|
135 |
"byte_fallback": true,
|
136 |
+
"ignore_merges": false,
|
137 |
"vocab": {
|
138 |
"<unk>": 0,
|
139 |
"<s>": 1,
|