Hugging Face's logo

alphaedge-ai
/
mmBERT-base-jav-16384

Fill-Mask
Transformers
Safetensors
Javanese
modernbert
feature-extraction
trimmed
Model card Files
xet
 Community
de-francophones commited on
Commit
aad26ad
·
verified ·
1 Parent(s): 499e1a6

Upload Javanese pruned model

Browse files
Files changed (5) hide show
  1. README.md +52 -0
  2. config.json +78 -0
  3. model.safetensors +3 -0
  4. tokenizer.json +0 -0
  5. tokenizer_config.json +17 -0
README.md ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: fill-mask
3
+ language: jav
4
+ license: mit
5
+ tags:
6
+ - trimmed
7
+ library_name: transformers
8
+ base_model: jhu-clsp/mmBERT-base
9
+ base_model_relation: quantized
10
+ datasets:
11
+ - Lumberjackk/fineweb-2-trimming
12
+ ---
13
+
14
+ # mmBERT-base-jav-32768
15
+ This model is a 55.86% smaller version of [jhu-clsp/mmBERT-base](https://huggingface.co/jhu-clsp/mmBERT-base) optimized for Javanese language via vocabulary size reduction using the [trimming](https://huggingface.co/blog/introduction-to-trimming) method.
16
+ This trimmed model should perform similarly to the original model with only 32,768 tokens and a much smaller memory footprint. However, it may not perform well for other languages as tokens not commonly used in the selected languages were removed from the vocabulary.
17
+
18
+ ## Model Statistics
19
+ | Metric | Original | Trimmed | Reduction |
20
+ |--------|----------|---------|-----------|
21
+ | **Vocabulary size** | 256,000 tokens | 32,768 tokens | **87.20%** |
22
+ | **Model size** | 306,939,648 params | 135,497,472 params | **55.86%** |
23
+
24
+ ![image](https://cdn-uploads.huggingface.co/production/uploads/613b0a62a14099d5afed7830/3bAHdqRvu-haO_RxyOwVo.png)
25
+
26
+ ## Mining Dataset Statistics
27
+ - **Number of texts used for mining**: 200,000 texts
28
+ - **Dataset**: [Lumberjackk/fineweb-2-trimming](https://huggingface.co/datasets/Lumberjackk/fineweb-2-trimming)
29
+
30
+ ## Usage
31
+ ```python
32
+ from transformers import AutoModel, AutoTokenizer
33
+
34
+ model_name = "Lumberjackk/mmBERT-base-jav-32768"
35
+ model = AutoModel.from_pretrained(model_name)
36
+ tokenizer = AutoTokenizer.from_pretrained(model_name)
37
+ ```
38
+
39
+ ## Citation
40
+
41
+ #### mmBERT
42
+ ```
43
+ @misc{marone2025mmbertmodernmultilingualencoder,
44
+ title={mmBERT: A Modern Multilingual Encoder with Annealed Language Learning},
45
+ author={Marc Marone and Orion Weller and William Fleshman and Eugene Yang and Dawn Lawrie and Benjamin Van Durme},
46
+ year={2025},
47
+ eprint={2509.06888},
48
+ archivePrefix={arXiv},
49
+ primaryClass={cs.CL},
50
+ url={https://arxiv.org/abs/2509.06888},
51
+ }
52
+ ```
config.json ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "ModernBertModel"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 2,
8
+ "classifier_activation": "gelu",
9
+ "classifier_bias": false,
10
+ "classifier_dropout": 0.0,
11
+ "classifier_pooling": "mean",
12
+ "cls_token_id": 1,
13
+ "decoder_bias": true,
14
+ "deterministic_flash_attn": false,
15
+ "dtype": "float32",
16
+ "embedding_dropout": 0.0,
17
+ "eos_token_id": 1,
18
+ "global_attn_every_n_layers": 3,
19
+ "gradient_checkpointing": false,
20
+ "hidden_activation": "gelu",
21
+ "hidden_size": 768,
22
+ "initializer_cutoff_factor": 2.0,
23
+ "initializer_range": 0.02,
24
+ "intermediate_size": 1152,
25
+ "layer_norm_eps": 1e-05,
26
+ "layer_types": [
27
+ "full_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "full_attention",
31
+ "sliding_attention",
32
+ "sliding_attention",
33
+ "full_attention",
34
+ "sliding_attention",
35
+ "sliding_attention",
36
+ "full_attention",
37
+ "sliding_attention",
38
+ "sliding_attention",
39
+ "full_attention",
40
+ "sliding_attention",
41
+ "sliding_attention",
42
+ "full_attention",
43
+ "sliding_attention",
44
+ "sliding_attention",
45
+ "full_attention",
46
+ "sliding_attention",
47
+ "sliding_attention",
48
+ "full_attention"
49
+ ],
50
+ "local_attention": 128,
51
+ "mask_token_id": 4,
52
+ "max_position_embeddings": 8192,
53
+ "mlp_bias": false,
54
+ "mlp_dropout": 0.0,
55
+ "model_type": "modernbert",
56
+ "norm_bias": false,
57
+ "norm_eps": 1e-05,
58
+ "num_attention_heads": 12,
59
+ "num_hidden_layers": 22,
60
+ "pad_token_id": 0,
61
+ "position_embedding_type": "sans_pos",
62
+ "rope_parameters": {
63
+ "full_attention": {
64
+ "rope_theta": 160000,
65
+ "rope_type": "default"
66
+ },
67
+ "sliding_attention": {
68
+ "rope_theta": 160000,
69
+ "rope_type": "default"
70
+ }
71
+ },
72
+ "sep_token_id": 1,
73
+ "sparse_pred_ignore_index": -100,
74
+ "sparse_prediction": false,
75
+ "tie_word_embeddings": true,
76
+ "transformers_version": "5.3.0.dev0",
77
+ "vocab_size": 16384
78
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40f97a3522f136af06757a57961b31111eae58c8acbb42ffdfc22db4a71b49ed
3
+ size 491671256
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "bos_token": "<bos>",
4
+ "cls_token": "<bos>",
5
+ "eos_token": "<eos>",
6
+ "mask_token": "<mask>",
7
+ "model_max_length": 8192,
8
+ "pad_token": "<pad>",
9
+ "padding_side": "right",
10
+ "sep_token": "<eos>",
11
+ "tokenizer_class": "TokenizersBackend",
12
+ "unk_token": "<unk>",
13
+ "model_input_names": [
14
+ "input_ids",
15
+ "attention_mask"
16
+ ]
17
+ }