Commit
•
5600af2
1
Parent(s):
f0b8bab
Push model using huggingface_hub.
Browse files- README.md +12 -0
- config.json +42 -0
- model.safetensors +3 -0
README.md
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
pipeline_tag: text_to_image
|
| 4 |
+
tags:
|
| 5 |
+
- arxiv:2304.12244
|
| 6 |
+
- model_hub_mixin
|
| 7 |
+
- pytorch_model_hub_mixin
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
|
| 11 |
+
- Library: [More Information Needed]
|
| 12 |
+
- Docs: [More Information Needed]
|
config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"experiment": {
|
| 3 |
+
"tokenizer_checkpoint": "tokenizer_titok_l32.bin",
|
| 4 |
+
"generator_checkpoint": "generator_titok_l32.bin",
|
| 5 |
+
"output_dir": "titok_l_32"
|
| 6 |
+
},
|
| 7 |
+
"model": {
|
| 8 |
+
"vq_model": {
|
| 9 |
+
"codebook_size": 4096,
|
| 10 |
+
"token_size": 12,
|
| 11 |
+
"use_l2_norm": true,
|
| 12 |
+
"commitment_cost": 0.25,
|
| 13 |
+
"vit_enc_model_size": "large",
|
| 14 |
+
"vit_dec_model_size": "large",
|
| 15 |
+
"vit_enc_patch_size": 16,
|
| 16 |
+
"vit_dec_patch_size": 16,
|
| 17 |
+
"num_latent_tokens": 32
|
| 18 |
+
},
|
| 19 |
+
"generator": {
|
| 20 |
+
"model_type": "ViT",
|
| 21 |
+
"hidden_size": 768,
|
| 22 |
+
"num_hidden_layers": 24,
|
| 23 |
+
"num_attention_heads": 16,
|
| 24 |
+
"intermediate_size": 3072,
|
| 25 |
+
"dropout": 0.1,
|
| 26 |
+
"attn_drop": 0.1,
|
| 27 |
+
"num_steps": 8,
|
| 28 |
+
"mask_schedule_strategy": "arccos",
|
| 29 |
+
"class_label_dropout": 0.1,
|
| 30 |
+
"image_seq_len": "${model.vq_model.num_latent_tokens}",
|
| 31 |
+
"condition_num_classes": 1000,
|
| 32 |
+
"randomize_temperature": 9.5,
|
| 33 |
+
"guidance_scale": 4.5,
|
| 34 |
+
"guidance_decay": "linear"
|
| 35 |
+
}
|
| 36 |
+
},
|
| 37 |
+
"dataset": {
|
| 38 |
+
"preprocessing": {
|
| 39 |
+
"crop_size": 256
|
| 40 |
+
}
|
| 41 |
+
}
|
| 42 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:832c5f1101e61d7e878b912e7ef0d6af43c39dfcf33c2b0b2f2924508b091a5b
|
| 3 |
+
size 708856136
|