Upload folder using huggingface_hub
Browse files- checkpoint-1004/config.json +56 -0
- checkpoint-1004/model.safetensors +3 -0
- checkpoint-1004/optimizer.pt +3 -0
- checkpoint-1004/preprocessor_config.json +24 -0
- checkpoint-1004/rng_state.pth +3 -0
- checkpoint-1004/scheduler.pt +3 -0
- checkpoint-1004/trainer_state.json +68 -0
- checkpoint-1004/training_args.bin +3 -0
- checkpoint-502/config.json +56 -0
- checkpoint-502/model.safetensors +3 -0
- checkpoint-502/optimizer.pt +3 -0
- checkpoint-502/preprocessor_config.json +24 -0
- checkpoint-502/rng_state.pth +3 -0
- checkpoint-502/scheduler.pt +3 -0
- checkpoint-502/trainer_state.json +51 -0
- checkpoint-502/training_args.bin +3 -0
- config.json +56 -0
- model.safetensors +3 -0
- preprocessor_config.json +24 -0
- training_args.bin +3 -0
checkpoint-1004/config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SiglipForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"id2label": {
|
| 6 |
+
"0": "Buildings and Structures",
|
| 7 |
+
"1": "Desert",
|
| 8 |
+
"2": "Forest Area",
|
| 9 |
+
"3": "Hill or Mountain",
|
| 10 |
+
"4": "Ice Glacier",
|
| 11 |
+
"5": "Sea or Ocean",
|
| 12 |
+
"6": "Street View"
|
| 13 |
+
},
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"label2id": {
|
| 16 |
+
"Buildings and Structures": 0,
|
| 17 |
+
"Desert": 1,
|
| 18 |
+
"Forest Area": 2,
|
| 19 |
+
"Hill or Mountain": 3,
|
| 20 |
+
"Ice Glacier": 4,
|
| 21 |
+
"Sea or Ocean": 5,
|
| 22 |
+
"Street View": 6
|
| 23 |
+
},
|
| 24 |
+
"model_type": "siglip",
|
| 25 |
+
"problem_type": "single_label_classification",
|
| 26 |
+
"text_config": {
|
| 27 |
+
"attention_dropout": 0.0,
|
| 28 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 29 |
+
"hidden_size": 768,
|
| 30 |
+
"intermediate_size": 3072,
|
| 31 |
+
"layer_norm_eps": 1e-06,
|
| 32 |
+
"max_position_embeddings": 64,
|
| 33 |
+
"model_type": "siglip_text_model",
|
| 34 |
+
"num_attention_heads": 12,
|
| 35 |
+
"num_hidden_layers": 12,
|
| 36 |
+
"projection_size": 768,
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"vocab_size": 256000
|
| 39 |
+
},
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.50.0",
|
| 42 |
+
"vision_config": {
|
| 43 |
+
"attention_dropout": 0.0,
|
| 44 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 45 |
+
"hidden_size": 768,
|
| 46 |
+
"image_size": 224,
|
| 47 |
+
"intermediate_size": 3072,
|
| 48 |
+
"layer_norm_eps": 1e-06,
|
| 49 |
+
"model_type": "siglip_vision_model",
|
| 50 |
+
"num_attention_heads": 12,
|
| 51 |
+
"num_channels": 3,
|
| 52 |
+
"num_hidden_layers": 12,
|
| 53 |
+
"patch_size": 16,
|
| 54 |
+
"torch_dtype": "float32"
|
| 55 |
+
}
|
| 56 |
+
}
|
checkpoint-1004/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:433992cdfa082e2ee54f3c418dcd7efde631d6cf4664d36e234bb1c431cc49f4
|
| 3 |
+
size 371583372
|
checkpoint-1004/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a953acbb5afab6e7262c1856bd136d6da0e2ea797173b742ed64d3a6c339c466
|
| 3 |
+
size 686586490
|
checkpoint-1004/preprocessor_config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "SiglipImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"processor_class": "SiglipProcessor",
|
| 18 |
+
"resample": 2,
|
| 19 |
+
"rescale_factor": 0.00392156862745098,
|
| 20 |
+
"size": {
|
| 21 |
+
"height": 224,
|
| 22 |
+
"width": 224
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-1004/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c40b0aacad198e52b153e48371141823bfad6760d269ffbeb53d4650fe06f051
|
| 3 |
+
size 14244
|
checkpoint-1004/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ebcea04e30e9945fbe1fbe0f610f17dbe785f48d2089969787c07028fd64711b
|
| 3 |
+
size 1064
|
checkpoint-1004/trainer_state.json
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1004,
|
| 3 |
+
"best_metric": 0.21446360647678375,
|
| 4 |
+
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-1004",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1004,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.9960159362549801,
|
| 14 |
+
"grad_norm": 2.575073003768921,
|
| 15 |
+
"learning_rate": 0.00010566037735849057,
|
| 16 |
+
"loss": 0.7675,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_accuracy": 0.7955466849622653,
|
| 22 |
+
"eval_loss": 0.6065428256988525,
|
| 23 |
+
"eval_model_preparation_time": 0.0023,
|
| 24 |
+
"eval_runtime": 206.8897,
|
| 25 |
+
"eval_samples_per_second": 77.495,
|
| 26 |
+
"eval_steps_per_second": 9.691,
|
| 27 |
+
"step": 502
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 1.9920318725099602,
|
| 31 |
+
"grad_norm": 4.37345552444458,
|
| 32 |
+
"learning_rate": 8.38574423480084e-07,
|
| 33 |
+
"loss": 0.37,
|
| 34 |
+
"step": 1000
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.0,
|
| 38 |
+
"eval_accuracy": 0.9245306555229839,
|
| 39 |
+
"eval_loss": 0.21446360647678375,
|
| 40 |
+
"eval_model_preparation_time": 0.0023,
|
| 41 |
+
"eval_runtime": 206.1909,
|
| 42 |
+
"eval_samples_per_second": 77.758,
|
| 43 |
+
"eval_steps_per_second": 9.724,
|
| 44 |
+
"step": 1004
|
| 45 |
+
}
|
| 46 |
+
],
|
| 47 |
+
"logging_steps": 500,
|
| 48 |
+
"max_steps": 1004,
|
| 49 |
+
"num_input_tokens_seen": 0,
|
| 50 |
+
"num_train_epochs": 2,
|
| 51 |
+
"save_steps": 500,
|
| 52 |
+
"stateful_callbacks": {
|
| 53 |
+
"TrainerControl": {
|
| 54 |
+
"args": {
|
| 55 |
+
"should_epoch_stop": false,
|
| 56 |
+
"should_evaluate": false,
|
| 57 |
+
"should_log": false,
|
| 58 |
+
"should_save": true,
|
| 59 |
+
"should_training_stop": true
|
| 60 |
+
},
|
| 61 |
+
"attributes": {}
|
| 62 |
+
}
|
| 63 |
+
},
|
| 64 |
+
"total_flos": 2.685815083993854e+18,
|
| 65 |
+
"train_batch_size": 32,
|
| 66 |
+
"trial_name": null,
|
| 67 |
+
"trial_params": null
|
| 68 |
+
}
|
checkpoint-1004/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12dd34e788f5eb534c4b6ac8f57ace245fefa2a536daa2fe035c88d0385d96e8
|
| 3 |
+
size 5304
|
checkpoint-502/config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SiglipForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"id2label": {
|
| 6 |
+
"0": "Buildings and Structures",
|
| 7 |
+
"1": "Desert",
|
| 8 |
+
"2": "Forest Area",
|
| 9 |
+
"3": "Hill or Mountain",
|
| 10 |
+
"4": "Ice Glacier",
|
| 11 |
+
"5": "Sea or Ocean",
|
| 12 |
+
"6": "Street View"
|
| 13 |
+
},
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"label2id": {
|
| 16 |
+
"Buildings and Structures": 0,
|
| 17 |
+
"Desert": 1,
|
| 18 |
+
"Forest Area": 2,
|
| 19 |
+
"Hill or Mountain": 3,
|
| 20 |
+
"Ice Glacier": 4,
|
| 21 |
+
"Sea or Ocean": 5,
|
| 22 |
+
"Street View": 6
|
| 23 |
+
},
|
| 24 |
+
"model_type": "siglip",
|
| 25 |
+
"problem_type": "single_label_classification",
|
| 26 |
+
"text_config": {
|
| 27 |
+
"attention_dropout": 0.0,
|
| 28 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 29 |
+
"hidden_size": 768,
|
| 30 |
+
"intermediate_size": 3072,
|
| 31 |
+
"layer_norm_eps": 1e-06,
|
| 32 |
+
"max_position_embeddings": 64,
|
| 33 |
+
"model_type": "siglip_text_model",
|
| 34 |
+
"num_attention_heads": 12,
|
| 35 |
+
"num_hidden_layers": 12,
|
| 36 |
+
"projection_size": 768,
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"vocab_size": 256000
|
| 39 |
+
},
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.50.0",
|
| 42 |
+
"vision_config": {
|
| 43 |
+
"attention_dropout": 0.0,
|
| 44 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 45 |
+
"hidden_size": 768,
|
| 46 |
+
"image_size": 224,
|
| 47 |
+
"intermediate_size": 3072,
|
| 48 |
+
"layer_norm_eps": 1e-06,
|
| 49 |
+
"model_type": "siglip_vision_model",
|
| 50 |
+
"num_attention_heads": 12,
|
| 51 |
+
"num_channels": 3,
|
| 52 |
+
"num_hidden_layers": 12,
|
| 53 |
+
"patch_size": 16,
|
| 54 |
+
"torch_dtype": "float32"
|
| 55 |
+
}
|
| 56 |
+
}
|
checkpoint-502/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:965730f18f6ba72d9aec55fca5bb19cddf16e334556357fa237ecc7ab53f218a
|
| 3 |
+
size 371583372
|
checkpoint-502/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:502bd51467ab759224edcdb26b212d22b5e39f96c49599a7e2d462765e9b2280
|
| 3 |
+
size 686586490
|
checkpoint-502/preprocessor_config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "SiglipImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"processor_class": "SiglipProcessor",
|
| 18 |
+
"resample": 2,
|
| 19 |
+
"rescale_factor": 0.00392156862745098,
|
| 20 |
+
"size": {
|
| 21 |
+
"height": 224,
|
| 22 |
+
"width": 224
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-502/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0873e1ed0482ff9b0f3522f2bb3be7fdddb965264ba2b9745b7fc59a08e5df6c
|
| 3 |
+
size 14244
|
checkpoint-502/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14a28031702231efca73f26a263fbb65de20124c7c88edf6eb948ec0ecc78c92
|
| 3 |
+
size 1064
|
checkpoint-502/trainer_state.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 502,
|
| 3 |
+
"best_metric": 0.6065428256988525,
|
| 4 |
+
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-502",
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 502,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.9960159362549801,
|
| 14 |
+
"grad_norm": 2.575073003768921,
|
| 15 |
+
"learning_rate": 0.00010566037735849057,
|
| 16 |
+
"loss": 0.7675,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_accuracy": 0.7955466849622653,
|
| 22 |
+
"eval_loss": 0.6065428256988525,
|
| 23 |
+
"eval_model_preparation_time": 0.0023,
|
| 24 |
+
"eval_runtime": 206.8897,
|
| 25 |
+
"eval_samples_per_second": 77.495,
|
| 26 |
+
"eval_steps_per_second": 9.691,
|
| 27 |
+
"step": 502
|
| 28 |
+
}
|
| 29 |
+
],
|
| 30 |
+
"logging_steps": 500,
|
| 31 |
+
"max_steps": 1004,
|
| 32 |
+
"num_input_tokens_seen": 0,
|
| 33 |
+
"num_train_epochs": 2,
|
| 34 |
+
"save_steps": 500,
|
| 35 |
+
"stateful_callbacks": {
|
| 36 |
+
"TrainerControl": {
|
| 37 |
+
"args": {
|
| 38 |
+
"should_epoch_stop": false,
|
| 39 |
+
"should_evaluate": false,
|
| 40 |
+
"should_log": false,
|
| 41 |
+
"should_save": true,
|
| 42 |
+
"should_training_stop": false
|
| 43 |
+
},
|
| 44 |
+
"attributes": {}
|
| 45 |
+
}
|
| 46 |
+
},
|
| 47 |
+
"total_flos": 1.342907541996927e+18,
|
| 48 |
+
"train_batch_size": 32,
|
| 49 |
+
"trial_name": null,
|
| 50 |
+
"trial_params": null
|
| 51 |
+
}
|
checkpoint-502/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12dd34e788f5eb534c4b6ac8f57ace245fefa2a536daa2fe035c88d0385d96e8
|
| 3 |
+
size 5304
|
config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"SiglipForImageClassification"
|
| 4 |
+
],
|
| 5 |
+
"id2label": {
|
| 6 |
+
"0": "Buildings and Structures",
|
| 7 |
+
"1": "Desert",
|
| 8 |
+
"2": "Forest Area",
|
| 9 |
+
"3": "Hill or Mountain",
|
| 10 |
+
"4": "Ice Glacier",
|
| 11 |
+
"5": "Sea or Ocean",
|
| 12 |
+
"6": "Street View"
|
| 13 |
+
},
|
| 14 |
+
"initializer_factor": 1.0,
|
| 15 |
+
"label2id": {
|
| 16 |
+
"Buildings and Structures": 0,
|
| 17 |
+
"Desert": 1,
|
| 18 |
+
"Forest Area": 2,
|
| 19 |
+
"Hill or Mountain": 3,
|
| 20 |
+
"Ice Glacier": 4,
|
| 21 |
+
"Sea or Ocean": 5,
|
| 22 |
+
"Street View": 6
|
| 23 |
+
},
|
| 24 |
+
"model_type": "siglip",
|
| 25 |
+
"problem_type": "single_label_classification",
|
| 26 |
+
"text_config": {
|
| 27 |
+
"attention_dropout": 0.0,
|
| 28 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 29 |
+
"hidden_size": 768,
|
| 30 |
+
"intermediate_size": 3072,
|
| 31 |
+
"layer_norm_eps": 1e-06,
|
| 32 |
+
"max_position_embeddings": 64,
|
| 33 |
+
"model_type": "siglip_text_model",
|
| 34 |
+
"num_attention_heads": 12,
|
| 35 |
+
"num_hidden_layers": 12,
|
| 36 |
+
"projection_size": 768,
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"vocab_size": 256000
|
| 39 |
+
},
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.50.0",
|
| 42 |
+
"vision_config": {
|
| 43 |
+
"attention_dropout": 0.0,
|
| 44 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 45 |
+
"hidden_size": 768,
|
| 46 |
+
"image_size": 224,
|
| 47 |
+
"intermediate_size": 3072,
|
| 48 |
+
"layer_norm_eps": 1e-06,
|
| 49 |
+
"model_type": "siglip_vision_model",
|
| 50 |
+
"num_attention_heads": 12,
|
| 51 |
+
"num_channels": 3,
|
| 52 |
+
"num_hidden_layers": 12,
|
| 53 |
+
"patch_size": 16,
|
| 54 |
+
"torch_dtype": "float32"
|
| 55 |
+
}
|
| 56 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:433992cdfa082e2ee54f3c418dcd7efde631d6cf4664d36e234bb1c431cc49f4
|
| 3 |
+
size 371583372
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_rgb": null,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_rescale": true,
|
| 5 |
+
"do_resize": true,
|
| 6 |
+
"image_mean": [
|
| 7 |
+
0.5,
|
| 8 |
+
0.5,
|
| 9 |
+
0.5
|
| 10 |
+
],
|
| 11 |
+
"image_processor_type": "SiglipImageProcessor",
|
| 12 |
+
"image_std": [
|
| 13 |
+
0.5,
|
| 14 |
+
0.5,
|
| 15 |
+
0.5
|
| 16 |
+
],
|
| 17 |
+
"processor_class": "SiglipProcessor",
|
| 18 |
+
"resample": 2,
|
| 19 |
+
"rescale_factor": 0.00392156862745098,
|
| 20 |
+
"size": {
|
| 21 |
+
"height": 224,
|
| 22 |
+
"width": 224
|
| 23 |
+
}
|
| 24 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12dd34e788f5eb534c4b6ac8f57ace245fefa2a536daa2fe035c88d0385d96e8
|
| 3 |
+
size 5304
|