Upload 12 files

assets.json

@@ -0,0 +1 @@
+{"class_name": "MultiModalPredictor", "column_types": {"image_path": "categorical", "xmax": "numerical", "xmin": "numerical", "ymax": "numerical", "ymin": "numerical", "image": "image_path", "name": "categorical"}, "label_column": "name", "problem_type": "multiclass", "presets": null, "eval_metric_name": "accuracy", "validation_metric_name": "accuracy", "output_shape": 10, "classes": null, "save_path": "C:\\Users\\iblfe\\Downloads\\dataset\\dataset\\AutogluonModels\\ag-20231031_101638", "pretrained": true, "pretrained_path": null, "fit_called": true, "best_score": null, "total_train_time": null, "version": "0.8.2"}

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "google/electra-base-discriminator",
+  "architectures": [
+    "ElectraForPreTraining"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "embedding_size": 768,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "electra",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "summary_activation": "gelu",
+  "summary_last_dropout": 0.1,
+  "summary_type": "first",
+  "summary_use_proj": true,
+  "transformers_version": "4.26.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

config.yaml

@@ -0,0 +1,147 @@
+model:
+  names:
+  - fusion_mlp
+  - hf_text
+  - numerical_mlp
+  - timm_image
+  numerical_mlp:
+    hidden_size: 128
+    activation: leaky_relu
+    num_layers: 1
+    drop_rate: 0.1
+    normalization: layer_norm
+    d_token: 8
+    embedding_arch: null
+    data_types:
+    - numerical
+    merge: concat
+  hf_text:
+    checkpoint_name: local://hf_text
+    gradient_checkpointing: false
+    pooling_mode: cls
+    data_types:
+    - text
+    tokenizer_name: hf_auto
+    max_text_len: 512
+    insert_sep: true
+    low_cpu_mem_usage: false
+    text_segment_num: 2
+    stochastic_chunk: false
+    text_aug_detect_length: 10
+    text_trivial_aug_maxscale: 0.0
+    text_train_augment_types: null
+  timm_image:
+    checkpoint_name: swin_base_patch4_window7_224
+    mix_choice: all_logits
+    data_types:
+    - image
+    train_transforms:
+    - resize_shorter_side
+    - center_crop
+    - trivial_augment
+    val_transforms:
+    - resize_shorter_side
+    - center_crop
+    image_norm: imagenet
+    image_size: null
+    max_img_num_per_col: 2
+  fusion_mlp:
+    weight: 0.1
+    adapt_in_features: max
+    hidden_sizes:
+    - 128
+    activation: leaky_relu
+    drop_rate: 0.1
+    normalization: layer_norm
+    data_types: null
+data:
+  image:
+    missing_value_strategy: zero
+  text:
+    normalize_text: false
+  categorical:
+    minimum_cat_count: 100
+    maximum_num_cat: 20
+    convert_to_text: true
+  numerical:
+    convert_to_text: false
+    scaler_with_mean: true
+    scaler_with_std: true
+  document:
+    missing_value_strategy: zero
+  label:
+    numerical_label_preprocessing: standardscaler
+  pos_label: null
+  mixup:
+    turn_on: false
+    mixup_alpha: 0.8
+    cutmix_alpha: 1.0
+    cutmix_minmax: null
+    prob: 1.0
+    switch_prob: 0.5
+    mode: batch
+    turn_off_epoch: 5
+    label_smoothing: 0.1
+  templates:
+    turn_on: false
+    num_templates: 30
+    template_length: 2048
+    preset_templates:
+    - super_glue
+    - rte
+    custom_templates: null
+optimization:
+  optim_type: adamw
+  learning_rate: 0.0001
+  weight_decay: 0.001
+  lr_choice: layerwise_decay
+  lr_decay: 0.9
+  lr_schedule: cosine_decay
+  max_epochs: 10
+  max_steps: -1
+  warmup_steps: 0.1
+  end_lr: 0
+  lr_mult: 1
+  patience: 10
+  val_check_interval: 0.5
+  check_val_every_n_epoch: 1
+  skip_final_val: false
+  gradient_clip_val: 1
+  gradient_clip_algorithm: norm
+  track_grad_norm: -1
+  log_every_n_steps: 10
+  top_k: 3
+  top_k_average_method: greedy_soup
+  efficient_finetune: null
+  lora:
+    module_filter: null
+    filter:
+    - query
+    - value
+    - ^q$
+    - ^v$
+    - ^k$
+    - ^o$
+    r: 8
+    alpha: 8
+  loss_function: auto
+  focal_loss:
+    alpha: null
+    gamma: 2.0
+    reduction: mean
+env:
+  num_gpus: 1
+  num_nodes: 1
+  batch_size: 128
+  per_gpu_batch_size: 8
+  eval_batch_size_ratio: 4
+  per_gpu_batch_size_evaluation: null
+  precision: 16
+  num_workers: 2
+  num_workers_evaluation: 2
+  fast_dev_run: false
+  deterministic: false
+  auto_select_gpus: true
+  strategy: null
+  deepspeed_allgather_size: 1000000000.0
+  deepspeed_allreduce_size: 1000000000.0

data_processors.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d176f6bba693fa5a50fde123e050dd2fa91e57296f738695bf3f30aa1ec4d9a2
+size 15164

df_preprocessor.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec63606a1d92a466866a4aa34fa8ea72b5f6e8a5fd12ab6616e5160ef749d957
+size 17162

events.out.tfevents.1698747405.影のリーダー.35680.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77f881a620c6840d8ecb273825dd5455a905936e80126a6f40f9957922415346
+size 3096003

hparams.yaml

@@ -0,0 +1,16 @@
+optim_type: adamw
+lr_choice: layerwise_decay
+lr_schedule: cosine_decay
+lr: 0.0001
+lr_decay: 0.9
+end_lr: 0
+lr_mult: 1
+weight_decay: 0.001
+warmup_steps: 0.1
+validation_metric_name: accuracy
+custom_metric_func: null
+efficient_finetune: null
+trainable_param_names: []
+mixup_fn: null
+mixup_off_epoch: 5
+skip_final_val: false

model.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa2b639b1d4d59750993081047b6f8d2802f35ac91a29e428694b63ec15dbe81
+size 792247793

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "name_or_path": "google/electra-base-discriminator",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "ElectraTokenizer",
+  "unk_token": "[UNK]"
+}