hoainam1706 commited on
Commit
1fa1501
·
verified ·
1 Parent(s): b53ac02

Delete UPLOAD_STEPS.md

Browse files
Files changed (1) hide show
  1. UPLOAD_STEPS.md +0 -63
UPLOAD_STEPS.md DELETED
@@ -1,63 +0,0 @@
1
- # Uploading WriteViT to Hugging Face
2
-
3
- This folder is a Hugging Face-ready copy of WriteViT with downloaded datasets/checkpoints and a polished card-style `README.md`.
4
-
5
- ## 1. Authenticate
6
-
7
- ```bash
8
- pip install -U huggingface_hub
9
- hf auth login
10
- ```
11
-
12
- ## 2. Choose Repository Type
13
-
14
- Because this release contains large dataset pickles and model checkpoints, a dataset repository is usually the cleanest choice:
15
-
16
- ```bash
17
- hf repo create YOUR_USERNAME/WriteViT --type dataset
18
- ```
19
-
20
- If you want it to appear under Hugging Face Models instead, use:
21
-
22
- ```bash
23
- hf repo create YOUR_USERNAME/WriteViT --type model
24
- ```
25
-
26
- ## 3. Upload Everything
27
-
28
- Dataset repo:
29
-
30
- ```bash
31
- cd /home/namhoai/WorkSpace/Hoctap/CS338/WriteViT-HF
32
- hf upload YOUR_USERNAME/WriteViT . --repo-type dataset
33
- ```
34
-
35
- Model repo:
36
-
37
- ```bash
38
- cd /home/namhoai/WorkSpace/Hoctap/CS338/WriteViT-HF
39
- hf upload YOUR_USERNAME/WriteViT . --repo-type model
40
- ```
41
-
42
- ## 4. Upload Only README Later
43
-
44
- If the large files are already uploaded and you only changed the card:
45
-
46
- ```bash
47
- hf upload YOUR_USERNAME/WriteViT README.md README.md --repo-type dataset
48
- ```
49
-
50
- Use `--repo-type model` if the existing repo is a model repo.
51
-
52
- ## Downloaded Artifacts
53
-
54
- ```text
55
- File/eng_ckpt.pth
56
- File/vn_ckpt.pth
57
- File/vgg19.pth
58
- File/IAM.pickle
59
- File/VN.pickle
60
- File/unifont.pickle
61
- File/english_words.txt
62
- File/vn_words.txt
63
- ```