Upload L4_uniform_distilled (distilled) with MTEB results

1_Pooling/config.json

@@ -0,0 +1,10 @@
+{
+    "word_embedding_dimension": 384,
+    "pooling_mode_cls_token": false,
+    "pooling_mode_mean_tokens": true,
+    "pooling_mode_max_tokens": false,
+    "pooling_mode_mean_sqrt_len_tokens": false,
+    "pooling_mode_weightedmean_tokens": false,
+    "pooling_mode_lasttoken": false,
+    "include_prompt": true
+}

README.md

@@ -0,0 +1,122 @@
+---
+language: ["ko", "en", "ja", "zh", "es", "fr", "de", "pt", "it", "ru", "ar", "hi", "th", "vi", "id", "tr", "nl", "pl"]
+tags:
+- sentence-transformers
+- intent-classification
+- multilingual
+- layer-pruning
+- vocab-pruning
+- knowledge-distillation
+library_name: sentence-transformers
+pipeline_tag: sentence-similarity
+license: apache-2.0
+---
+
+# L4_uniform_distilled (Distilled)
+
+Lightweight multilingual sentence encoder optimized for intent classification.
+Created from `paraphrase-multilingual-MiniLM-L12-v2` via layer pruning + corpus-based vocabulary pruning + knowledge distillation.
+
+## Model Details
+
+| Property | Value |
+|----------|-------|
+| Teacher | paraphrase-multilingual-MiniLM-L12-v2 |
+| Architecture | XLM-RoBERTa (pruned) |
+| Hidden dim | 384 |
+| Layers | 4 / 12 |
+| Layer indices | [0, 4, 7, 11] |
+| Strategy | 4 layers, evenly spaced (compact) |
+| Vocab size | ~38,330 (pruned from 250K) |
+| Parameters | 22,642,560 |
+| Safetensors size | 84.6MB |
+| Distilled | Yes |
+
+## Supported Languages (18)
+
+ko, en, ja, zh, es, fr, de, pt, it, ru, ar, hi, th, vi, id, tr, nl, pl
+
+## Quick Start
+
+```python
+from sentence_transformers import SentenceTransformer
+
+model = SentenceTransformer("L4_uniform_distilled")
+
+sentences = [
+    "예약 좀 해줘",           # Korean
+    "What did I order?",     # English
+    "今日はいい天気ですね",    # Japanese
+    "Reserva una mesa",      # Spanish
+]
+
+embeddings = model.encode(sentences)
+print(embeddings.shape)  # (4, 384)
+```
+
+## MTEB Evaluation Results
+
+**Overall Average: 54.61%**
+
+### MassiveIntentClassification
+
+**Average: 50.88%**
+
+| Language | Score |
+|----------|-------|
+| ar | 41.25% |
+| en | 62.02% |
+| es | 53.15% |
+| ko | 47.08% |
+
+### MassiveScenarioClassification
+
+**Average: 58.34%**
+
+| Language | Score |
+|----------|-------|
+| ar | 47.4% |
+| en | 69.96% |
+| es | 61.24% |
+| ko | 54.74% |
+
+
+## Distillation Impact
+
+| Task | Before Distillation | After Distillation | Delta |
+|------|--------------------|--------------------|-------|
+| MassiveIntentClassification | 50.25% | 50.88% | +0.63%p |
+| MassiveScenarioClassification | 53.82% | 58.34% | +4.52%p |
+
+
+## Training
+
+This model was created in two stages:
+
+### Stage 1: Layer Pruning
+1. Teacher model: `paraphrase-multilingual-MiniLM-L12-v2` (12 layers, 384 hidden dim)
+2. Selected layers: `[0, 4, 7, 11]` (4 layers, evenly spaced (compact))
+3. Vocabulary pruning: 250K -> ~38K tokens (corpus-based, 18 target languages)
+
+### Stage 2: Knowledge Distillation
+- **Method**: MSE + Cosine Similarity loss between teacher and student embeddings
+- **Training data**: MASSIVE dataset (90K multilingual sentences, 18 languages)
+- **Optimizer**: AdamW (lr=2e-5, weight_decay=0.01)
+- **Schedule**: Cosine annealing over 3 epochs
+- **Batch size**: 64
+- **Base model**: `L4_uniform` (layer-pruned only)
+
+
+## Compression Summary
+
+| Stage | Vocab | Layers | Size |
+|-------|-------|--------|------|
+| Teacher (original) | 250,002 | 12 | ~480MB |
+| + Layer pruning | 250,002 | 4 | ~393MB |
+| + Vocab pruning | ~38,330 | 4 | ~85MB |
+
+## Limitations
+
+- Vocabulary pruning restricts the model to the 18 target languages
+- Designed for short dialogue utterances, not long documents
+- Layer pruning may reduce performance on complex semantic tasks

config.json

@@ -0,0 +1,25 @@
+{
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 384,
+  "initializer_range": 0.02,
+  "intermediate_size": 1536,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 4,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.56.2",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 38330
+}

config_sentence_transformers.json

@@ -0,0 +1,14 @@
+{
+  "model_type": "SentenceTransformer",
+  "__version__": {
+    "sentence_transformers": "5.3.0",
+    "transformers": "4.56.2",
+    "pytorch": "2.10.0+cu128"
+  },
+  "prompts": {
+    "query": "",
+    "document": ""
+  },
+  "default_prompt_name": null,
+  "similarity_fn_name": "cosine"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:849314db7a40fefcafc751585b9a2004108365166b922a06d5c979451da025e4
+size 88658024

modules.json

@@ -0,0 +1,14 @@
+[
+  {
+    "idx": 0,
+    "name": "0",
+    "path": "",
+    "type": "sentence_transformers.models.Transformer"
+  },
+  {
+    "idx": 1,
+    "name": "1",
+    "path": "1_Pooling",
+    "type": "sentence_transformers.models.Pooling"
+  }
+]

sentence_bert_config.json

@@ -0,0 +1,4 @@
+{
+    "max_seq_length": 128,
+    "do_lower_case": false
+}

special_tokens_map.json

@@ -0,0 +1,51 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "cls_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "mask_token": {
+    "content": "<mask>",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "sep_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,65 @@
+{
+  "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
+    },
+    "38329": {
+      "content": "<mask>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<s>",
+  "do_lower_case": true,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "mask_token": "<mask>",
+  "max_length": 128,
+  "model_max_length": 128,
+  "pad_to_multiple_of": null,
+  "pad_token": "<pad>",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
+  "sep_token": "</s>",
+  "stride": 0,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "<unk>"
+}