Laya fine-tuned on GIZ vulnerability data: macro-F1 0.000 (zero-shot 0.000)

README.md

@@ -0,0 +1,46 @@
+---
+license: apache-2.0
+library_name: transformers
+tags: [laya, text-classification, multi-label, climate, vulnerability, rlcd]
+pipeline_tag: text-classification
+---
+
+# Laya fine-tuned for climate-vulnerability group detection (multi-label)
+
+[convaiinnovations/laya](https://huggingface.co/convaiinnovations/laya) (421M, ModernBERT-large backbone)
+fine-tuned with the official RLCD recipe on [GIZ/vulnerability_training_data_full](https://huggingface.co/datasets/GIZ/vulnerability_training_data_full)
+(380 train rows x 17 binary vulnerability-group questions; 36 all-negative rows included as negatives).
+
+Each vulnerability group is asked as one binary (`noul`) typed question; all 17 are answered in a single forward pass.
+Evaluate with `laya.load("peter2000/laya-vulnerability-groups")` and `agent.predict(state, questions)`.
+
+## Test-set metrics (held-out 5 rows, threshold 0.5)
+
+| metric | value |
+|---|---|
+| macro-F1 | 0.0000 |
+| micro-F1 | 0.0000 |
+| ECE | 0.0688 |
+| subset accuracy | 0.0000 |
+
+Per-label F1:
+
+| label | F1 |
+|---|---|
+| Agricultural communities | 0.0000 |
+| Coastal communities | 0.0000 |
+| Ethnic, racial or other minorities | 0.0000 |
+| Fishery communities | 0.0000 |
+| Informal sector workers | 0.0000 |
+| Members of indigenous and local communities | 0.0000 |
+| Migrants and displaced persons | 0.0000 |
+| Older persons | 0.0000 |
+| Other | 0.0000 |
+| Persons living in poverty | 0.0000 |
+| Persons with disabilities | 0.0000 |
+| Persons with pre-existing health conditions | 0.0000 |
+| Residents of drought-prone regions | 0.0000 |
+| Rural populations | 0.0000 |
+| Sexual minorities (LGBTQI+) | 0.0000 |
+| Urban populations | 0.0000 |
+| Women and other genders | 0.0000 |

encoder/config.json

@@ -0,0 +1,84 @@
+{
+  "architectures": [
+    "ModernBertForMaskedLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "dtype": "float32",
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 1024,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 2624,
+  "layer_norm_eps": 1e-05,
+  "layer_types": [
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "local_attention": 128,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 28,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "repad_logits_with_grad": false,
+  "rope_parameters": {
+    "full_attention": {
+      "rope_theta": 160000.0,
+      "rope_type": "default"
+    },
+    "sliding_attention": {
+      "rope_theta": 10000.0,
+      "rope_type": "default"
+    }
+  },
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.17.0",
+  "vocab_size": 50368
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a87929b1f9599ac659d97a89662a6f8806ff5ae15ad1c1e27e89e55ae24187e4
+size 842609220

rl_agent_config.json

@@ -0,0 +1,35 @@
+{
+  "encoder": "answerdotai/ModernBERT-large",
+  "head_layers": 2,
+  "max_len": 512,
+  "head_max_len": 192,
+  "max_prefixes": 6,
+  "act_costs": {
+    "escalate": 0.5
+  },
+  "cost_wrong_act": 3.0,
+  "amp_dtype": "bf16",
+  "model_name": "laya-vulnerability-groups",
+  "temperature": [
+    1.2,
+    1.2,
+    1.0
+  ],
+  "temperature_by_options": {
+    "choice:3-5": 1.7601518630981445,
+    "choice:6-10": 1.0000158548355103,
+    "score:3-5": 1.2514300346374512,
+    "noul:2": 1.0,
+    "choice:11+": 0.10058280825614929,
+    "choice:2": 1.9063563346862793
+  },
+  "training": {
+    "updates": 7313,
+    "epochs_completed": 1,
+    "hours": 1.96,
+    "world_size": 1,
+    "fine_tuned_from_checkpoint": true
+  },
+  "gradient_checkpointing": true,
+  "fine_tuned": true
+}

tokenizer/tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "local_files_only": false,
+  "mask_token": "[MASK]",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 8192,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "[UNK]"
+}