first commit

This commit is contained in:
yang yingjie 2025-05-30 14:08:12 +08:00
parent 6d674315bd
commit 86a8679b6c
28 changed files with 573 additions and 0 deletions

View File

@ -0,0 +1,52 @@
{
"architectures": [
"SiglipForImageClassification"
],
"id2label": {
"0": "Anime Picture",
"1": "Hentai",
"2": "Normal",
"3": "Pornography",
"4": "Enticing or Sensual"
},
"initializer_factor": 1.0,
"label2id": {
"Anime Picture": 0,
"Enticing or Sensual": 4,
"Hentai": 1,
"Normal": 2,
"Pornography": 3
},
"model_type": "siglip",
"problem_type": "single_label_classification",
"text_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"max_position_embeddings": 64,
"model_type": "siglip_text_model",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"projection_size": 768,
"torch_dtype": "float32",
"vocab_size": 256000
},
"torch_dtype": "float32",
"transformers_version": "4.50.0",
"vision_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"image_size": 256,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"model_type": "siglip_vision_model",
"num_attention_heads": 12,
"num_channels": 3,
"num_hidden_layers": 12,
"patch_size": 16,
"torch_dtype": "float32"
}
}

BIN
checkpoint-1656/model.safetensors (Stored with Git LFS) Normal file

Binary file not shown.

BIN
checkpoint-1656/optimizer.pt (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,24 @@
{
"do_convert_rgb": null,
"do_normalize": true,
"do_rescale": true,
"do_resize": true,
"image_mean": [
0.5,
0.5,
0.5
],
"image_processor_type": "SiglipImageProcessor",
"image_std": [
0.5,
0.5,
0.5
],
"processor_class": "SiglipProcessor",
"resample": 2,
"rescale_factor": 0.00392156862745098,
"size": {
"height": 256,
"width": 256
}
}

BIN
checkpoint-1656/rng_state.pth (Stored with Git LFS) Normal file

Binary file not shown.

BIN
checkpoint-1656/scheduler.pt (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,75 @@
{
"best_global_step": 1656,
"best_metric": 0.33448827266693115,
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-1656",
"epoch": 2.0,
"eval_steps": 500,
"global_step": 1656,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.6038647342995169,
"grad_norm": 5.860389709472656,
"learning_rate": 0.00016302382908792113,
"loss": 0.823,
"step": 500
},
{
"epoch": 1.0,
"eval_accuracy": 0.7698523581165276,
"eval_loss": 0.577109158039093,
"eval_model_preparation_time": 0.0023,
"eval_runtime": 605.3087,
"eval_samples_per_second": 43.751,
"eval_steps_per_second": 5.47,
"step": 828
},
{
"epoch": 1.2077294685990339,
"grad_norm": 5.253721237182617,
"learning_rate": 0.00012193919474116682,
"loss": 0.6183,
"step": 1000
},
{
"epoch": 1.8115942028985508,
"grad_norm": 4.610513210296631,
"learning_rate": 8.08545603944125e-05,
"loss": 0.5136,
"step": 1500
},
{
"epoch": 2.0,
"eval_accuracy": 0.8734282369822151,
"eval_loss": 0.33448827266693115,
"eval_model_preparation_time": 0.0023,
"eval_runtime": 598.3057,
"eval_samples_per_second": 44.263,
"eval_steps_per_second": 5.534,
"step": 1656
}
],
"logging_steps": 500,
"max_steps": 2484,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.794356182313075e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}

BIN
checkpoint-1656/training_args.bin (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,52 @@
{
"architectures": [
"SiglipForImageClassification"
],
"id2label": {
"0": "Anime Picture",
"1": "Hentai",
"2": "Normal",
"3": "Pornography",
"4": "Enticing or Sensual"
},
"initializer_factor": 1.0,
"label2id": {
"Anime Picture": 0,
"Enticing or Sensual": 4,
"Hentai": 1,
"Normal": 2,
"Pornography": 3
},
"model_type": "siglip",
"problem_type": "single_label_classification",
"text_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"max_position_embeddings": 64,
"model_type": "siglip_text_model",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"projection_size": 768,
"torch_dtype": "float32",
"vocab_size": 256000
},
"torch_dtype": "float32",
"transformers_version": "4.50.0",
"vision_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"image_size": 256,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"model_type": "siglip_vision_model",
"num_attention_heads": 12,
"num_channels": 3,
"num_hidden_layers": 12,
"patch_size": 16,
"torch_dtype": "float32"
}
}

BIN
checkpoint-2484/model.safetensors (Stored with Git LFS) Normal file

Binary file not shown.

BIN
checkpoint-2484/optimizer.pt (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,24 @@
{
"do_convert_rgb": null,
"do_normalize": true,
"do_rescale": true,
"do_resize": true,
"image_mean": [
0.5,
0.5,
0.5
],
"image_processor_type": "SiglipImageProcessor",
"image_std": [
0.5,
0.5,
0.5
],
"processor_class": "SiglipProcessor",
"resample": 2,
"rescale_factor": 0.00392156862745098,
"size": {
"height": 256,
"width": 256
}
}

BIN
checkpoint-2484/rng_state.pth (Stored with Git LFS) Normal file

Binary file not shown.

BIN
checkpoint-2484/scheduler.pt (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,92 @@
{
"best_global_step": 2484,
"best_metric": 0.23217608034610748,
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-2484",
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2484,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.6038647342995169,
"grad_norm": 5.860389709472656,
"learning_rate": 0.00016302382908792113,
"loss": 0.823,
"step": 500
},
{
"epoch": 1.0,
"eval_accuracy": 0.7698523581165276,
"eval_loss": 0.577109158039093,
"eval_model_preparation_time": 0.0023,
"eval_runtime": 605.3087,
"eval_samples_per_second": 43.751,
"eval_steps_per_second": 5.47,
"step": 828
},
{
"epoch": 1.2077294685990339,
"grad_norm": 5.253721237182617,
"learning_rate": 0.00012193919474116682,
"loss": 0.6183,
"step": 1000
},
{
"epoch": 1.8115942028985508,
"grad_norm": 4.610513210296631,
"learning_rate": 8.08545603944125e-05,
"loss": 0.5136,
"step": 1500
},
{
"epoch": 2.0,
"eval_accuracy": 0.8734282369822151,
"eval_loss": 0.33448827266693115,
"eval_model_preparation_time": 0.0023,
"eval_runtime": 598.3057,
"eval_samples_per_second": 44.263,
"eval_steps_per_second": 5.534,
"step": 1656
},
{
"epoch": 2.4154589371980677,
"grad_norm": 5.658606052398682,
"learning_rate": 3.976992604765818e-05,
"loss": 0.3894,
"step": 2000
},
{
"epoch": 3.0,
"eval_accuracy": 0.9137182343390099,
"eval_loss": 0.23217608034610748,
"eval_model_preparation_time": 0.0023,
"eval_runtime": 604.5474,
"eval_samples_per_second": 43.806,
"eval_steps_per_second": 5.477,
"step": 2484
}
],
"logging_steps": 500,
"max_steps": 2484,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 8.691534273469612e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}

BIN
checkpoint-2484/training_args.bin (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,52 @@
{
"architectures": [
"SiglipForImageClassification"
],
"id2label": {
"0": "Anime Picture",
"1": "Hentai",
"2": "Normal",
"3": "Pornography",
"4": "Enticing or Sensual"
},
"initializer_factor": 1.0,
"label2id": {
"Anime Picture": 0,
"Enticing or Sensual": 4,
"Hentai": 1,
"Normal": 2,
"Pornography": 3
},
"model_type": "siglip",
"problem_type": "single_label_classification",
"text_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"max_position_embeddings": 64,
"model_type": "siglip_text_model",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"projection_size": 768,
"torch_dtype": "float32",
"vocab_size": 256000
},
"torch_dtype": "float32",
"transformers_version": "4.50.0",
"vision_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"image_size": 256,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"model_type": "siglip_vision_model",
"num_attention_heads": 12,
"num_channels": 3,
"num_hidden_layers": 12,
"patch_size": 16,
"torch_dtype": "float32"
}
}

BIN
checkpoint-828/model.safetensors (Stored with Git LFS) Normal file

Binary file not shown.

BIN
checkpoint-828/optimizer.pt (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,24 @@
{
"do_convert_rgb": null,
"do_normalize": true,
"do_rescale": true,
"do_resize": true,
"image_mean": [
0.5,
0.5,
0.5
],
"image_processor_type": "SiglipImageProcessor",
"image_std": [
0.5,
0.5,
0.5
],
"processor_class": "SiglipProcessor",
"resample": 2,
"rescale_factor": 0.00392156862745098,
"size": {
"height": 256,
"width": 256
}
}

BIN
checkpoint-828/rng_state.pth (Stored with Git LFS) Normal file

Binary file not shown.

BIN
checkpoint-828/scheduler.pt (Stored with Git LFS) Normal file

Binary file not shown.

View File

@ -0,0 +1,51 @@
{
"best_global_step": 828,
"best_metric": 0.577109158039093,
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-828",
"epoch": 1.0,
"eval_steps": 500,
"global_step": 828,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.6038647342995169,
"grad_norm": 5.860389709472656,
"learning_rate": 0.00016302382908792113,
"loss": 0.823,
"step": 500
},
{
"epoch": 1.0,
"eval_accuracy": 0.7698523581165276,
"eval_loss": 0.577109158039093,
"eval_model_preparation_time": 0.0023,
"eval_runtime": 605.3087,
"eval_samples_per_second": 43.751,
"eval_steps_per_second": 5.47,
"step": 828
}
],
"logging_steps": 500,
"max_steps": 2484,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.8971780911565373e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}

BIN
checkpoint-828/training_args.bin (Stored with Git LFS) Normal file

Binary file not shown.

52
config.json Normal file
View File

@ -0,0 +1,52 @@
{
"architectures": [
"SiglipForImageClassification"
],
"id2label": {
"0": "Anime Picture",
"1": "Hentai",
"2": "Normal",
"3": "Pornography",
"4": "Enticing or Sensual"
},
"initializer_factor": 1.0,
"label2id": {
"Anime Picture": 0,
"Enticing or Sensual": 4,
"Hentai": 1,
"Normal": 2,
"Pornography": 3
},
"model_type": "siglip",
"problem_type": "single_label_classification",
"text_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"max_position_embeddings": 64,
"model_type": "siglip_text_model",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"projection_size": 768,
"torch_dtype": "float32",
"vocab_size": 256000
},
"torch_dtype": "float32",
"transformers_version": "4.50.0",
"vision_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"image_size": 256,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"model_type": "siglip_vision_model",
"num_attention_heads": 12,
"num_channels": 3,
"num_hidden_layers": 12,
"patch_size": 16,
"torch_dtype": "float32"
}
}

BIN
model.safetensors (Stored with Git LFS) Normal file

Binary file not shown.

24
preprocessor_config.json Normal file
View File

@ -0,0 +1,24 @@
{
"do_convert_rgb": null,
"do_normalize": true,
"do_rescale": true,
"do_resize": true,
"image_mean": [
0.5,
0.5,
0.5
],
"image_processor_type": "SiglipImageProcessor",
"image_std": [
0.5,
0.5,
0.5
],
"processor_class": "SiglipProcessor",
"resample": 2,
"rescale_factor": 0.00392156862745098,
"size": {
"height": 256,
"width": 256
}
}

BIN
training_args.bin (Stored with Git LFS) Normal file

Binary file not shown.