Upload folder using huggingface_hub
Browse files- checkpoint-1062/config.json +50 -0
- checkpoint-1062/model.safetensors +3 -0
- checkpoint-1062/optimizer.pt +3 -0
- checkpoint-1062/preprocessor_config.json +24 -0
- checkpoint-1062/rng_state.pth +3 -0
- checkpoint-1062/scheduler.pt +3 -0
- checkpoint-1062/trainer_state.json +58 -0
- checkpoint-1062/training_args.bin +3 -0
- checkpoint-2124/config.json +50 -0
- checkpoint-2124/model.safetensors +3 -0
- checkpoint-2124/optimizer.pt +3 -0
- checkpoint-2124/preprocessor_config.json +24 -0
- checkpoint-2124/rng_state.pth +3 -0
- checkpoint-2124/scheduler.pt +3 -0
- checkpoint-2124/trainer_state.json +82 -0
- checkpoint-2124/training_args.bin +3 -0
- checkpoint-3186/config.json +50 -0
- checkpoint-3186/model.safetensors +3 -0
- checkpoint-3186/optimizer.pt +3 -0
- checkpoint-3186/preprocessor_config.json +24 -0
- checkpoint-3186/rng_state.pth +3 -0
- checkpoint-3186/scheduler.pt +3 -0
- checkpoint-3186/trainer_state.json +106 -0
- checkpoint-3186/training_args.bin +3 -0
- config.json +50 -0
- model.safetensors +3 -0
- preprocessor_config.json +24 -0
- training_args.bin +3 -0
checkpoint-1062/config.json
ADDED
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"SiglipForImageClassification"
|
4 |
+
],
|
5 |
+
"id2label": {
|
6 |
+
"0": "MildDemented",
|
7 |
+
"1": "ModerateDemented",
|
8 |
+
"2": "NonDemented",
|
9 |
+
"3": "VeryMildDemented"
|
10 |
+
},
|
11 |
+
"initializer_factor": 1.0,
|
12 |
+
"label2id": {
|
13 |
+
"MildDemented": 0,
|
14 |
+
"ModerateDemented": 1,
|
15 |
+
"NonDemented": 2,
|
16 |
+
"VeryMildDemented": 3
|
17 |
+
},
|
18 |
+
"model_type": "siglip",
|
19 |
+
"problem_type": "single_label_classification",
|
20 |
+
"text_config": {
|
21 |
+
"attention_dropout": 0.0,
|
22 |
+
"hidden_act": "gelu_pytorch_tanh",
|
23 |
+
"hidden_size": 768,
|
24 |
+
"intermediate_size": 3072,
|
25 |
+
"layer_norm_eps": 1e-06,
|
26 |
+
"max_position_embeddings": 64,
|
27 |
+
"model_type": "siglip_text_model",
|
28 |
+
"num_attention_heads": 12,
|
29 |
+
"num_hidden_layers": 12,
|
30 |
+
"projection_size": 768,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"vocab_size": 256000
|
33 |
+
},
|
34 |
+
"torch_dtype": "float32",
|
35 |
+
"transformers_version": "4.50.0",
|
36 |
+
"vision_config": {
|
37 |
+
"attention_dropout": 0.0,
|
38 |
+
"hidden_act": "gelu_pytorch_tanh",
|
39 |
+
"hidden_size": 768,
|
40 |
+
"image_size": 224,
|
41 |
+
"intermediate_size": 3072,
|
42 |
+
"layer_norm_eps": 1e-06,
|
43 |
+
"model_type": "siglip_vision_model",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_channels": 3,
|
46 |
+
"num_hidden_layers": 12,
|
47 |
+
"patch_size": 16,
|
48 |
+
"torch_dtype": "float32"
|
49 |
+
}
|
50 |
+
}
|
checkpoint-1062/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:24141abda9e9309e74aa8a4ddae8ce8238d1752730f77978004103b5948bcbd1
|
3 |
+
size 371574144
|
checkpoint-1062/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a66ff7697d5a444f1cfd80ec60ca9fc39cb6a021210eab14fc709fb3497d4888
|
3 |
+
size 686568058
|
checkpoint-1062/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-1062/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6684114d55602e59d9c5110b849969368bb32b3548cb50eaddd4934c50a8ce11
|
3 |
+
size 14244
|
checkpoint-1062/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:efccceda213691294969464fa824bd4fd6e51df5322f0c20f9f5dca76c80eaed
|
3 |
+
size 1064
|
checkpoint-1062/trainer_state.json
ADDED
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": 1062,
|
3 |
+
"best_metric": 0.7104782462120056,
|
4 |
+
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-1062",
|
5 |
+
"epoch": 1.0,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 1062,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.4708097928436911,
|
14 |
+
"grad_norm": 10.774321556091309,
|
15 |
+
"learning_rate": 8.565051020408164e-05,
|
16 |
+
"loss": 1.047,
|
17 |
+
"step": 500
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.9416195856873822,
|
21 |
+
"grad_norm": 4.252574920654297,
|
22 |
+
"learning_rate": 6.970663265306123e-05,
|
23 |
+
"loss": 0.7732,
|
24 |
+
"step": 1000
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 1.0,
|
28 |
+
"eval_accuracy": 0.6644597457627118,
|
29 |
+
"eval_loss": 0.7104782462120056,
|
30 |
+
"eval_model_preparation_time": 0.0038,
|
31 |
+
"eval_runtime": 419.7148,
|
32 |
+
"eval_samples_per_second": 80.969,
|
33 |
+
"eval_steps_per_second": 10.121,
|
34 |
+
"step": 1062
|
35 |
+
}
|
36 |
+
],
|
37 |
+
"logging_steps": 500,
|
38 |
+
"max_steps": 3186,
|
39 |
+
"num_input_tokens_seen": 0,
|
40 |
+
"num_train_epochs": 3,
|
41 |
+
"save_steps": 500,
|
42 |
+
"stateful_callbacks": {
|
43 |
+
"TrainerControl": {
|
44 |
+
"args": {
|
45 |
+
"should_epoch_stop": false,
|
46 |
+
"should_evaluate": false,
|
47 |
+
"should_log": false,
|
48 |
+
"should_save": true,
|
49 |
+
"should_training_stop": false
|
50 |
+
},
|
51 |
+
"attributes": {}
|
52 |
+
}
|
53 |
+
},
|
54 |
+
"total_flos": 2.846393976227365e+18,
|
55 |
+
"train_batch_size": 32,
|
56 |
+
"trial_name": null,
|
57 |
+
"trial_params": null
|
58 |
+
}
|
checkpoint-1062/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e61a621ea5a8d382bc2e56cb7a8ae6a3ad4ed62aa0cbc4ea8610f5957a76edd2
|
3 |
+
size 5304
|
checkpoint-2124/config.json
ADDED
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"SiglipForImageClassification"
|
4 |
+
],
|
5 |
+
"id2label": {
|
6 |
+
"0": "MildDemented",
|
7 |
+
"1": "ModerateDemented",
|
8 |
+
"2": "NonDemented",
|
9 |
+
"3": "VeryMildDemented"
|
10 |
+
},
|
11 |
+
"initializer_factor": 1.0,
|
12 |
+
"label2id": {
|
13 |
+
"MildDemented": 0,
|
14 |
+
"ModerateDemented": 1,
|
15 |
+
"NonDemented": 2,
|
16 |
+
"VeryMildDemented": 3
|
17 |
+
},
|
18 |
+
"model_type": "siglip",
|
19 |
+
"problem_type": "single_label_classification",
|
20 |
+
"text_config": {
|
21 |
+
"attention_dropout": 0.0,
|
22 |
+
"hidden_act": "gelu_pytorch_tanh",
|
23 |
+
"hidden_size": 768,
|
24 |
+
"intermediate_size": 3072,
|
25 |
+
"layer_norm_eps": 1e-06,
|
26 |
+
"max_position_embeddings": 64,
|
27 |
+
"model_type": "siglip_text_model",
|
28 |
+
"num_attention_heads": 12,
|
29 |
+
"num_hidden_layers": 12,
|
30 |
+
"projection_size": 768,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"vocab_size": 256000
|
33 |
+
},
|
34 |
+
"torch_dtype": "float32",
|
35 |
+
"transformers_version": "4.50.0",
|
36 |
+
"vision_config": {
|
37 |
+
"attention_dropout": 0.0,
|
38 |
+
"hidden_act": "gelu_pytorch_tanh",
|
39 |
+
"hidden_size": 768,
|
40 |
+
"image_size": 224,
|
41 |
+
"intermediate_size": 3072,
|
42 |
+
"layer_norm_eps": 1e-06,
|
43 |
+
"model_type": "siglip_vision_model",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_channels": 3,
|
46 |
+
"num_hidden_layers": 12,
|
47 |
+
"patch_size": 16,
|
48 |
+
"torch_dtype": "float32"
|
49 |
+
}
|
50 |
+
}
|
checkpoint-2124/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b58eef82e75e6d874b2ddb886dc114d78259d87d1085843d9e308c7fd011f27b
|
3 |
+
size 371574144
|
checkpoint-2124/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:65f6396a8aaf558ca64fa52db7950300d1c165bcd4b2753ce70e4c6e258b06aa
|
3 |
+
size 686568058
|
checkpoint-2124/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-2124/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3c1b345df5c81bda9ba90af5a37d39677a8e7e02dc1bedd401cb985646cf096e
|
3 |
+
size 14244
|
checkpoint-2124/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ee0e312a8ee32a1c41c4ffb32ea4f99f7801ccc87e21176a55a879adf377a462
|
3 |
+
size 1064
|
checkpoint-2124/trainer_state.json
ADDED
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": 2124,
|
3 |
+
"best_metric": 0.43832501769065857,
|
4 |
+
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-2124",
|
5 |
+
"epoch": 2.0,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 2124,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.4708097928436911,
|
14 |
+
"grad_norm": 10.774321556091309,
|
15 |
+
"learning_rate": 8.565051020408164e-05,
|
16 |
+
"loss": 1.047,
|
17 |
+
"step": 500
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.9416195856873822,
|
21 |
+
"grad_norm": 4.252574920654297,
|
22 |
+
"learning_rate": 6.970663265306123e-05,
|
23 |
+
"loss": 0.7732,
|
24 |
+
"step": 1000
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 1.0,
|
28 |
+
"eval_accuracy": 0.6644597457627118,
|
29 |
+
"eval_loss": 0.7104782462120056,
|
30 |
+
"eval_model_preparation_time": 0.0038,
|
31 |
+
"eval_runtime": 419.7148,
|
32 |
+
"eval_samples_per_second": 80.969,
|
33 |
+
"eval_steps_per_second": 10.121,
|
34 |
+
"step": 1062
|
35 |
+
},
|
36 |
+
{
|
37 |
+
"epoch": 1.4124293785310735,
|
38 |
+
"grad_norm": 4.366853713989258,
|
39 |
+
"learning_rate": 5.3762755102040816e-05,
|
40 |
+
"loss": 0.6655,
|
41 |
+
"step": 1500
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"epoch": 1.8832391713747647,
|
45 |
+
"grad_norm": 7.878406047821045,
|
46 |
+
"learning_rate": 3.781887755102041e-05,
|
47 |
+
"loss": 0.5723,
|
48 |
+
"step": 2000
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 2.0,
|
52 |
+
"eval_accuracy": 0.8096162900188324,
|
53 |
+
"eval_loss": 0.43832501769065857,
|
54 |
+
"eval_model_preparation_time": 0.0038,
|
55 |
+
"eval_runtime": 419.9197,
|
56 |
+
"eval_samples_per_second": 80.93,
|
57 |
+
"eval_steps_per_second": 10.116,
|
58 |
+
"step": 2124
|
59 |
+
}
|
60 |
+
],
|
61 |
+
"logging_steps": 500,
|
62 |
+
"max_steps": 3186,
|
63 |
+
"num_input_tokens_seen": 0,
|
64 |
+
"num_train_epochs": 3,
|
65 |
+
"save_steps": 500,
|
66 |
+
"stateful_callbacks": {
|
67 |
+
"TrainerControl": {
|
68 |
+
"args": {
|
69 |
+
"should_epoch_stop": false,
|
70 |
+
"should_evaluate": false,
|
71 |
+
"should_log": false,
|
72 |
+
"should_save": true,
|
73 |
+
"should_training_stop": false
|
74 |
+
},
|
75 |
+
"attributes": {}
|
76 |
+
}
|
77 |
+
},
|
78 |
+
"total_flos": 5.69278795245473e+18,
|
79 |
+
"train_batch_size": 32,
|
80 |
+
"trial_name": null,
|
81 |
+
"trial_params": null
|
82 |
+
}
|
checkpoint-2124/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e61a621ea5a8d382bc2e56cb7a8ae6a3ad4ed62aa0cbc4ea8610f5957a76edd2
|
3 |
+
size 5304
|
checkpoint-3186/config.json
ADDED
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"SiglipForImageClassification"
|
4 |
+
],
|
5 |
+
"id2label": {
|
6 |
+
"0": "MildDemented",
|
7 |
+
"1": "ModerateDemented",
|
8 |
+
"2": "NonDemented",
|
9 |
+
"3": "VeryMildDemented"
|
10 |
+
},
|
11 |
+
"initializer_factor": 1.0,
|
12 |
+
"label2id": {
|
13 |
+
"MildDemented": 0,
|
14 |
+
"ModerateDemented": 1,
|
15 |
+
"NonDemented": 2,
|
16 |
+
"VeryMildDemented": 3
|
17 |
+
},
|
18 |
+
"model_type": "siglip",
|
19 |
+
"problem_type": "single_label_classification",
|
20 |
+
"text_config": {
|
21 |
+
"attention_dropout": 0.0,
|
22 |
+
"hidden_act": "gelu_pytorch_tanh",
|
23 |
+
"hidden_size": 768,
|
24 |
+
"intermediate_size": 3072,
|
25 |
+
"layer_norm_eps": 1e-06,
|
26 |
+
"max_position_embeddings": 64,
|
27 |
+
"model_type": "siglip_text_model",
|
28 |
+
"num_attention_heads": 12,
|
29 |
+
"num_hidden_layers": 12,
|
30 |
+
"projection_size": 768,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"vocab_size": 256000
|
33 |
+
},
|
34 |
+
"torch_dtype": "float32",
|
35 |
+
"transformers_version": "4.50.0",
|
36 |
+
"vision_config": {
|
37 |
+
"attention_dropout": 0.0,
|
38 |
+
"hidden_act": "gelu_pytorch_tanh",
|
39 |
+
"hidden_size": 768,
|
40 |
+
"image_size": 224,
|
41 |
+
"intermediate_size": 3072,
|
42 |
+
"layer_norm_eps": 1e-06,
|
43 |
+
"model_type": "siglip_vision_model",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_channels": 3,
|
46 |
+
"num_hidden_layers": 12,
|
47 |
+
"patch_size": 16,
|
48 |
+
"torch_dtype": "float32"
|
49 |
+
}
|
50 |
+
}
|
checkpoint-3186/model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8e3b1807bac6c56faeac282de76f583ac281ed74cb00f30935e5c3be5bbe5b82
|
3 |
+
size 371574144
|
checkpoint-3186/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:732726181667eccb3ee8355b06664453d6036df572a5605dbcc814f2b2402a8f
|
3 |
+
size 686568058
|
checkpoint-3186/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-3186/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:feb341863255fa652ebcc3f277791628409694cb003399b86783f3db005e3e6f
|
3 |
+
size 14244
|
checkpoint-3186/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:70acd8f7be020d79e7e7a56668ba5a57ae6e7b0ad13ff439c5b1615b572c6729
|
3 |
+
size 1064
|
checkpoint-3186/trainer_state.json
ADDED
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_global_step": 3186,
|
3 |
+
"best_metric": 0.1701197475194931,
|
4 |
+
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-3186",
|
5 |
+
"epoch": 3.0,
|
6 |
+
"eval_steps": 500,
|
7 |
+
"global_step": 3186,
|
8 |
+
"is_hyper_param_search": false,
|
9 |
+
"is_local_process_zero": true,
|
10 |
+
"is_world_process_zero": true,
|
11 |
+
"log_history": [
|
12 |
+
{
|
13 |
+
"epoch": 0.4708097928436911,
|
14 |
+
"grad_norm": 10.774321556091309,
|
15 |
+
"learning_rate": 8.565051020408164e-05,
|
16 |
+
"loss": 1.047,
|
17 |
+
"step": 500
|
18 |
+
},
|
19 |
+
{
|
20 |
+
"epoch": 0.9416195856873822,
|
21 |
+
"grad_norm": 4.252574920654297,
|
22 |
+
"learning_rate": 6.970663265306123e-05,
|
23 |
+
"loss": 0.7732,
|
24 |
+
"step": 1000
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 1.0,
|
28 |
+
"eval_accuracy": 0.6644597457627118,
|
29 |
+
"eval_loss": 0.7104782462120056,
|
30 |
+
"eval_model_preparation_time": 0.0038,
|
31 |
+
"eval_runtime": 419.7148,
|
32 |
+
"eval_samples_per_second": 80.969,
|
33 |
+
"eval_steps_per_second": 10.121,
|
34 |
+
"step": 1062
|
35 |
+
},
|
36 |
+
{
|
37 |
+
"epoch": 1.4124293785310735,
|
38 |
+
"grad_norm": 4.366853713989258,
|
39 |
+
"learning_rate": 5.3762755102040816e-05,
|
40 |
+
"loss": 0.6655,
|
41 |
+
"step": 1500
|
42 |
+
},
|
43 |
+
{
|
44 |
+
"epoch": 1.8832391713747647,
|
45 |
+
"grad_norm": 7.878406047821045,
|
46 |
+
"learning_rate": 3.781887755102041e-05,
|
47 |
+
"loss": 0.5723,
|
48 |
+
"step": 2000
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 2.0,
|
52 |
+
"eval_accuracy": 0.8096162900188324,
|
53 |
+
"eval_loss": 0.43832501769065857,
|
54 |
+
"eval_model_preparation_time": 0.0038,
|
55 |
+
"eval_runtime": 419.9197,
|
56 |
+
"eval_samples_per_second": 80.93,
|
57 |
+
"eval_steps_per_second": 10.116,
|
58 |
+
"step": 2124
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 2.354048964218456,
|
62 |
+
"grad_norm": 11.523202896118164,
|
63 |
+
"learning_rate": 2.1875e-05,
|
64 |
+
"loss": 0.4148,
|
65 |
+
"step": 2500
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 2.824858757062147,
|
69 |
+
"grad_norm": 4.949053764343262,
|
70 |
+
"learning_rate": 5.931122448979593e-06,
|
71 |
+
"loss": 0.2886,
|
72 |
+
"step": 3000
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 3.0,
|
76 |
+
"eval_accuracy": 0.931379472693032,
|
77 |
+
"eval_loss": 0.1701197475194931,
|
78 |
+
"eval_model_preparation_time": 0.0038,
|
79 |
+
"eval_runtime": 415.1862,
|
80 |
+
"eval_samples_per_second": 81.852,
|
81 |
+
"eval_steps_per_second": 10.232,
|
82 |
+
"step": 3186
|
83 |
+
}
|
84 |
+
],
|
85 |
+
"logging_steps": 500,
|
86 |
+
"max_steps": 3186,
|
87 |
+
"num_input_tokens_seen": 0,
|
88 |
+
"num_train_epochs": 3,
|
89 |
+
"save_steps": 500,
|
90 |
+
"stateful_callbacks": {
|
91 |
+
"TrainerControl": {
|
92 |
+
"args": {
|
93 |
+
"should_epoch_stop": false,
|
94 |
+
"should_evaluate": false,
|
95 |
+
"should_log": false,
|
96 |
+
"should_save": true,
|
97 |
+
"should_training_stop": true
|
98 |
+
},
|
99 |
+
"attributes": {}
|
100 |
+
}
|
101 |
+
},
|
102 |
+
"total_flos": 8.539181928682095e+18,
|
103 |
+
"train_batch_size": 32,
|
104 |
+
"trial_name": null,
|
105 |
+
"trial_params": null
|
106 |
+
}
|
checkpoint-3186/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e61a621ea5a8d382bc2e56cb7a8ae6a3ad4ed62aa0cbc4ea8610f5957a76edd2
|
3 |
+
size 5304
|
config.json
ADDED
@@ -0,0 +1,50 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"architectures": [
|
3 |
+
"SiglipForImageClassification"
|
4 |
+
],
|
5 |
+
"id2label": {
|
6 |
+
"0": "MildDemented",
|
7 |
+
"1": "ModerateDemented",
|
8 |
+
"2": "NonDemented",
|
9 |
+
"3": "VeryMildDemented"
|
10 |
+
},
|
11 |
+
"initializer_factor": 1.0,
|
12 |
+
"label2id": {
|
13 |
+
"MildDemented": 0,
|
14 |
+
"ModerateDemented": 1,
|
15 |
+
"NonDemented": 2,
|
16 |
+
"VeryMildDemented": 3
|
17 |
+
},
|
18 |
+
"model_type": "siglip",
|
19 |
+
"problem_type": "single_label_classification",
|
20 |
+
"text_config": {
|
21 |
+
"attention_dropout": 0.0,
|
22 |
+
"hidden_act": "gelu_pytorch_tanh",
|
23 |
+
"hidden_size": 768,
|
24 |
+
"intermediate_size": 3072,
|
25 |
+
"layer_norm_eps": 1e-06,
|
26 |
+
"max_position_embeddings": 64,
|
27 |
+
"model_type": "siglip_text_model",
|
28 |
+
"num_attention_heads": 12,
|
29 |
+
"num_hidden_layers": 12,
|
30 |
+
"projection_size": 768,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"vocab_size": 256000
|
33 |
+
},
|
34 |
+
"torch_dtype": "float32",
|
35 |
+
"transformers_version": "4.50.0",
|
36 |
+
"vision_config": {
|
37 |
+
"attention_dropout": 0.0,
|
38 |
+
"hidden_act": "gelu_pytorch_tanh",
|
39 |
+
"hidden_size": 768,
|
40 |
+
"image_size": 224,
|
41 |
+
"intermediate_size": 3072,
|
42 |
+
"layer_norm_eps": 1e-06,
|
43 |
+
"model_type": "siglip_vision_model",
|
44 |
+
"num_attention_heads": 12,
|
45 |
+
"num_channels": 3,
|
46 |
+
"num_hidden_layers": 12,
|
47 |
+
"patch_size": 16,
|
48 |
+
"torch_dtype": "float32"
|
49 |
+
}
|
50 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8e3b1807bac6c56faeac282de76f583ac281ed74cb00f30935e5c3be5bbe5b82
|
3 |
+
size 371574144
|
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:e61a621ea5a8d382bc2e56cb7a8ae6a3ad4ed62aa0cbc4ea8610f5957a76edd2
|
3 |
+
size 5304
|