initial setupt

2025-08-06 22:45:37 +01:00
commit fef3f5ae35
42 changed files with 7147 additions and 0 deletions
@@ -0,0 +1,30 @@
+task:
+  name: "matching"
+  type: "semantic_matching"
+
+data:
+  source: "huggingface"
+  dataset_name: "sentence-transformers/paraphrase-MiniLM-L3-v2"
+  input_field: "sentence1"
+  target_field: "sentence2"
+  label_field: "label"
+  max_length: 128
+  train_split: 0.8
+  validation_split: 0.1
+  test_split: 0.1
+
+model:
+  name: "sentence-transformers/all-MiniLM-L6-v2"
+  max_length: 128
+
+training:
+  num_epochs: 3
+  batch_size: 32
+  learning_rate: 2e-5
+  weight_decay: 0.01
+  warmup_ratio: 0.1
+  lr_scheduler_type: "linear"
+
+inference:
+  batch_size: 64
+  similarity_threshold: 0.5