Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+optuna.db filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,32 @@
+---
+language: vi
+license: cc-by-nc-sa-4.0
+library_name: pytorch
+tags:
+- esg
+- phobert
+- multi-label
+- vietnamese
+- text-classification
+---
+
+# huypham71/esg-topic
+
+PhoBERT multi-head (sigmoid) cho ESG — backbone `vinai/phobert-base`, các đầu: `['env', 'soc', 'gov']`.
+Đây là kiến trúc **tùy biến** (một encoder + nhiều đầu nhị phân độc lập, masked BCE),
+KHÔNG phải `AutoModelForSequenceClassification`, nên cần dùng code `esgwash` để tải.
+
+- **Thresholds theo từng đầu** (tune trên tập val, tối đa hóa F1): `{'env': 0.27, 'soc': 0.13, 'gov': 0.46}`
+- File: `model.safetensors` (backbone + các đầu), tokenizer PhoBERT, `config.json`.
+
+## Cách tải
+```python
+import json
+from esgwash.models.trainer import MultiHeadTrainer
+from huggingface_hub import snapshot_download
+
+d = snapshot_download("huypham71/esg-topic")
+cfg = json.load(open(f"{d}/config.json"))
+model = MultiHeadTrainer(cfg).load(d)
+probs = model.predict_proba(["câu tiếng Việt ..."])  # tự word-segment bên trong
+```

added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "<mask>": 64000
+}

best_params.json

@@ -0,0 +1,13 @@
+{
+  "params": {
+    "lr": 2.7396836004694616e-05,
+    "batch_size": 16,
+    "warmup_ratio": 0.06467163677523774,
+    "weight_decay": 0.019448553647326953,
+    "dropout": 0.05249460840378278
+  },
+  "val_macro_f1": 0.8571,
+  "epochs": 10,
+  "n_trials": 20,
+  "tuned_at": "2026-06-13T10:58:20"
+}

config.json

@@ -0,0 +1,74 @@
+{
+  "backbone": "vinai/phobert-base",
+  "en_backbone": "roberta-base",
+  "zero_shot_backbone": "xlm-roberta-base",
+  "max_length": 256,
+  "word_segment": true,
+  "labels": [
+    "env",
+    "soc",
+    "gov"
+  ],
+  "loss": "masked_bce",
+  "pos_weight": "auto",
+  "train": {
+    "lr": 2.7396836004694616e-05,
+    "batch_size": 16,
+    "epochs": 10,
+    "warmup_ratio": 0.06467163677523774,
+    "weight_decay": 0.019448553647326953,
+    "seeds": [
+      42,
+      43,
+      44,
+      45,
+      46
+    ]
+  },
+  "threshold_tuning": "per_label_f1",
+  "tune": {
+    "max_epochs": 10,
+    "space": {
+      "lr": [
+        1e-05,
+        5e-05
+      ],
+      "batch_size": [
+        16,
+        32
+      ],
+      "warmup_ratio": [
+        0.0,
+        0.2
+      ],
+      "weight_decay": [
+        0.001,
+        0.1
+      ],
+      "dropout": [
+        0.05,
+        0.3
+      ]
+    }
+  },
+  "cross_label": {
+    "confidence": 0.9,
+    "probs_path": "data/processed/gold/esgbert_probs.parquet",
+    "table_path": "data/processed/gold/topic_labeled.parquet"
+  },
+  "data": {
+    "masked_table": "data/processed/gold/topic_masked.parquet"
+  },
+  "output_dir": "outputs/models/topic",
+  "heads": [
+    "env",
+    "soc",
+    "gov"
+  ],
+  "dropout": 0.05249460840378278,
+  "thresholds": {
+    "env": 0.27,
+    "soc": 0.13,
+    "gov": 0.46
+  }
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2912e27e8e08d0bf7066a948afefa53a38db7c5cd93186e7ff925a43a1c2872d
+size 540026964

optuna.db

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aadd65a05defec1415e4457fcc245e884bfed1be61c0979253b77c7796ea2f8f
+size 131072

tokenizer_config.json

@@ -0,0 +1,56 @@
+{
+  "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
+    },
+    "64000": {
+      "content": "<mask>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "backend": "custom",
+  "bos_token": "<s>",
+  "cls_token": "<s>",
+  "eos_token": "</s>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "sep_token": "</s>",
+  "tokenizer_class": "PhobertTokenizer",
+  "unk_token": "<unk>"
+}