Upload folder using huggingface_hub

assets.json

@@ -0,0 +1 @@
+{"class_name": "BaseLearner", "column_types": {"image_path": "image_path", "text": "text", "label": "categorical"}, "label_column": "label", "problem_type": "binary", "presets": null, "eval_metric_name": "accuracy", "validation_metric_name": "accuracy", "minmax_mode": "max", "output_shape": 2, "save_path": "/content/AutogluonModels/ag-20240108_103835", "pretrained": true, "pretrained_path": null, "fit_called": true, "best_score": null, "total_train_time": null, "version": "1.0.0"}

config.yaml

@@ -0,0 +1,148 @@
+model:
+  names:
+  - fusion_mlp
+  - hf_text
+  - timm_image
+  hf_text:
+    checkpoint_name: local://hf_text
+    gradient_checkpointing: false
+    pooling_mode: cls
+    data_types:
+    - text
+    tokenizer_name: hf_auto
+    use_fast: true
+    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: timm/swinv2_base_window8_256.ms_in1k
+    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
+  column_features_pooling_mode: concat
+  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
+  mask2former_loss:
+    loss_cross_entropy_weight: 10.0
+    loss_mask_weight: 5.0
+    loss_dice_weight: 5.0
+  extra_trainable_params: []
+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-mixed
+  num_workers: 2
+  num_workers_evaluation: 2
+  accelerator: auto
+  fast_dev_run: false
+  deterministic: false
+  auto_select_gpus: true
+  strategy: auto
+  deepspeed_allgather_size: 1000000000.0
+  deepspeed_allreduce_size: 1000000000.0
+  compile:
+    turn_on: false
+    mode: default
+    dynamic: true
+    backend: inductor

data_processors.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:043bcc7958ac9e1fe6eb535f2c0fdeda746066f01a84d14b8f7e81f95e2c42fc
+size 15421

df_preprocessor.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a9c5ef7dacf348ac78db5ae5a939c7993375c9806559a9d490b32b6a5ba8a96
+size 14093

events.out.tfevents.1704710458.acc5ce4afb75.1271.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:291ef20f12da6eb0e11098b1c6046e4c4c33e8f7a50453e0b5d927c3cf2d3759
+size 857275

hf_text/config.json

@@ -0,0 +1,48 @@
+{
+  "_name_or_path": "cardiffnlp/twitter-roberta-large-hate-latest",
+  "architectures": [
+    "RobertaModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 0,
+  "classifier_dropout": null,
+  "eos_token_id": 2,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "hate_gender",
+    "1": "hate_race",
+    "2": "hate_sexuality",
+    "3": "hate_religion",
+    "4": "hate_origin",
+    "5": "hate_disability",
+    "6": "hate_age",
+    "7": "not_hate"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "label2id": {
+    "hate_age": 6,
+    "hate_disability": 5,
+    "hate_gender": 0,
+    "hate_origin": 4,
+    "hate_race": 1,
+    "hate_religion": 3,
+    "hate_sexuality": 2,
+    "not_hate": 7
+  },
+  "layer_norm_eps": 1e-05,
+  "max_position_embeddings": 514,
+  "model_type": "roberta",
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "pad_token_id": 1,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.31.0",
+  "type_vocab_size": 1,
+  "use_cache": true,
+  "vocab_size": 50265
+}

hf_text/special_tokens_map.json

@@ -0,0 +1,15 @@
+{
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "unk_token": "<unk>"
+}

hf_text/tokenizer_config.json

@@ -0,0 +1,61 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "50264": {
+      "content": "<mask>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "errors": "replace",
+  "mask_token": "<mask>",
+  "max_length": 64,
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "stride": 0,
+  "tokenizer_class": "RobertaTokenizer",
+  "trim_offsets": true,
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "<unk>"
+}

hparams.yaml

@@ -0,0 +1,15 @@
+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
+mixup_off_epoch: 5
+skip_final_val: false
+track_grad_norm: -1

model.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a69917c07f5bfccdbdc920beb15c3ffc70861b233897eb7b7282b6ec7e2929b
+size 1778714829