config
Browse files- README.md +9 -0
- config.json +88 -0
README.md
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- pytorch_model_hub_mixin
|
| 4 |
+
- model_hub_mixin
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
|
| 8 |
+
- Library: [More Information Needed]
|
| 9 |
+
- Docs: [More Information Needed]
|
config.json
ADDED
|
@@ -0,0 +1,88 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cfg": {
|
| 3 |
+
"base_lr": 0.0001,
|
| 4 |
+
"batch_size": 256,
|
| 5 |
+
"batch_size_val": 256,
|
| 6 |
+
"beta1": 0.5,
|
| 7 |
+
"beta2": 0.5,
|
| 8 |
+
"ckpt_path": "/data02/xy/GeoRSCLIP/codebase/inference/pretrain/RS5M_ViT-B-32.pt",
|
| 9 |
+
"clip_pretrain": "./cisen/pretrain/RS5M_ViT-B-32.pt",
|
| 10 |
+
"contrastive": [
|
| 11 |
+
0.4,
|
| 12 |
+
0.3,
|
| 13 |
+
0.3
|
| 14 |
+
],
|
| 15 |
+
"dataset": "classification",
|
| 16 |
+
"dataset_json_file": "/data02/xy/dataEngine/json_data/LuojiaHOG(test)_.json",
|
| 17 |
+
"dataset_query_split": 0.2,
|
| 18 |
+
"dataset_train_split": 0.6,
|
| 19 |
+
"dim_ffn": 2048,
|
| 20 |
+
"dist_backend": "nccl",
|
| 21 |
+
"dist_url": "tcp://localhost:3693",
|
| 22 |
+
"dropout": 0.1,
|
| 23 |
+
"emb_dim": 768,
|
| 24 |
+
"epochs": 50,
|
| 25 |
+
"eta": 0.2,
|
| 26 |
+
"evaluate": true,
|
| 27 |
+
"exp_name": "classifi",
|
| 28 |
+
"fpn_in": [
|
| 29 |
+
512,
|
| 30 |
+
768,
|
| 31 |
+
768
|
| 32 |
+
],
|
| 33 |
+
"fpn_out": [
|
| 34 |
+
768,
|
| 35 |
+
768,
|
| 36 |
+
768,
|
| 37 |
+
512
|
| 38 |
+
],
|
| 39 |
+
"gpu": [
|
| 40 |
+
0
|
| 41 |
+
],
|
| 42 |
+
"heads": 12,
|
| 43 |
+
"image_resolution": 224,
|
| 44 |
+
"imgs_folder": "/data02/xy/Clip-hash/datasets/image/",
|
| 45 |
+
"input_size": 224,
|
| 46 |
+
"intermediate": false,
|
| 47 |
+
"label_path": "/data02/xy/Clip-hash/labels.txt",
|
| 48 |
+
"lamda1": 0.5,
|
| 49 |
+
"lamda2": 0.5,
|
| 50 |
+
"lr_decay": 0.5,
|
| 51 |
+
"lr_multi": 0.1,
|
| 52 |
+
"manual_seed": 0,
|
| 53 |
+
"max_norm": 0.0,
|
| 54 |
+
"milestones": [
|
| 55 |
+
50
|
| 56 |
+
],
|
| 57 |
+
"min_lr": 1e-08,
|
| 58 |
+
"model_name": "ViT-B-32",
|
| 59 |
+
"multiprocessing_distributed": true,
|
| 60 |
+
"num_classes": 10,
|
| 61 |
+
"num_head": 8,
|
| 62 |
+
"num_layers": 3,
|
| 63 |
+
"output_dim": 512,
|
| 64 |
+
"output_folder": "/data02/xy/Clip-hash/exp/",
|
| 65 |
+
"print_freq": 1,
|
| 66 |
+
"rank": 0,
|
| 67 |
+
"ratio": 0,
|
| 68 |
+
"resume": false,
|
| 69 |
+
"save_freq": 1,
|
| 70 |
+
"start_epoch": 0,
|
| 71 |
+
"sync_bn": true,
|
| 72 |
+
"test_batch_size": 256,
|
| 73 |
+
"test_lmdb": "/data02/xy/Clip-hash/datasets/lmdb/refcoco/val.lmdb",
|
| 74 |
+
"test_split": "val-test",
|
| 75 |
+
"topk": 5,
|
| 76 |
+
"val_batch_size": 1,
|
| 77 |
+
"vis_dim": 512,
|
| 78 |
+
"visualize": false,
|
| 79 |
+
"warmup_epochs": 0,
|
| 80 |
+
"weight": null,
|
| 81 |
+
"weight_decay": 0.0,
|
| 82 |
+
"word_dim": 1024,
|
| 83 |
+
"word_len": 328,
|
| 84 |
+
"workers": 32,
|
| 85 |
+
"workers_val": 16,
|
| 86 |
+
"world_size": 1
|
| 87 |
+
}
|
| 88 |
+
}
|