Hugging Face's logo

codechrl
/
bert-micro-cybersecurity

Fill-Mask
Transformers
Safetensors
PyTorch
English
Indonesian
bert
text-classification
token-classification
cybersecurity
named-entity-recognition
tensorflow
masked-language-modeling
Model card Files
xet
 Community
codechrl commited on
Commit
8d88f71
·
verified ·
1 Parent(s): e86507e

Training update: 18,961/238,664 rows (7.94%) | +3361 new @ 2025-10-23 16:13:17

Browse files
Files changed (4) hide show
  1. README.md +5 -5
  2. model.safetensors +1 -1
  3. training_args.bin +1 -1
  4. training_metadata.json +7 -7
README.md CHANGED
@@ -25,7 +25,7 @@ pipeline_tag: fill-mask
25
  - Model type: fine-tuned lightweight BERT variant
26
  - Languages: English & Indonesia
27
  - Finetuned from: `boltuix/bert-micro`
28
- - Status: **Early version** — trained on **7.64%** of planned data.
29
 
30
  **Model sources**
31
  - Base model: [boltuix/bert-micro](https://huggingface.co/boltuix/bert-micro)
@@ -51,7 +51,7 @@ You can use this model to classify cybersecurity-related text — for example, w
51
  - Early classification of SIEM alert & events.
52
 
53
  ## 3. Bias, Risks, and Limitations
54
- Because the model is based on a small subset (7.64%) of planned data, performance is preliminary and may degrade on unseen or specialized domains (industrial control, IoT logs, foreign language).
55
  - Inherits any biases present in the base model (`boltuix/bert-micro`) and in the fine-tuning data — e.g., over-representation of certain threat types, vendor or tooling-specific vocabulary.
56
  - **Should not be used as sole authority for incident decisions; only as an aid to human analysts.**
57
 
@@ -75,9 +75,9 @@ Since cybersecurity data often contains lengthy alert descriptions and execution
75
  - **LR scheduler**: Linear with warmup
76
 
77
  ### Training Data
78
- - **Total database rows**: 238,638
79
- - **Rows processed (cumulative)**: 18,225 (7.64%)
80
- - **Training date**: 2025-10-23 13:59:44
81
 
82
  ### Post-Training Metrics
83
  - **Final training loss**:
 
25
  - Model type: fine-tuned lightweight BERT variant
26
  - Languages: English & Indonesia
27
  - Finetuned from: `boltuix/bert-micro`
28
+ - Status: **Early version** — trained on **7.94%** of planned data.
29
 
30
  **Model sources**
31
  - Base model: [boltuix/bert-micro](https://huggingface.co/boltuix/bert-micro)
 
51
  - Early classification of SIEM alert & events.
52
 
53
  ## 3. Bias, Risks, and Limitations
54
+ Because the model is based on a small subset (7.94%) of planned data, performance is preliminary and may degrade on unseen or specialized domains (industrial control, IoT logs, foreign language).
55
  - Inherits any biases present in the base model (`boltuix/bert-micro`) and in the fine-tuning data — e.g., over-representation of certain threat types, vendor or tooling-specific vocabulary.
56
  - **Should not be used as sole authority for incident decisions; only as an aid to human analysts.**
57
 
 
75
  - **LR scheduler**: Linear with warmup
76
 
77
  ### Training Data
78
+ - **Total database rows**: 238,664
79
+ - **Rows processed (cumulative)**: 18,961 (7.94%)
80
+ - **Training date**: 2025-10-23 16:13:17
81
 
82
  ### Post-Training Metrics
83
  - **Final training loss**:
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b6b179f72defc538fd7b5653127462bf075d20412d34ca869e94fc8e59df157c
3
  size 17671560
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ba4e82f04b18adaec539499e048d543bbb946d628685bfa5593e27885c3872f
3
  size 17671560
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4252431a1f0cf6586c1e1401c68e8bc871c544cf3223ef1c8c486934847a4cb8
3
  size 5905
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a42b400bb65959b60751bd8fd3dd9553968f73bd20d0837f6d25ac0ad7e8713
3
  size 5905
training_metadata.json CHANGED
@@ -1,11 +1,11 @@
1
  {
2
- "trained_at": 1761227984.984249,
3
- "trained_at_readable": "2025-10-23 13:59:44",
4
- "samples_this_session": 4997,
5
- "new_rows_this_session": 742,
6
- "trained_rows_total": 18225,
7
- "total_db_rows": 238638,
8
- "percentage": 7.6370904885223645,
9
  "final_loss": 0,
10
  "epochs": 3,
11
  "learning_rate": 5e-05,
 
1
  {
2
+ "trained_at": 1761235997.13213,
3
+ "trained_at_readable": "2025-10-23 16:13:17",
4
+ "samples_this_session": 4992,
5
+ "new_rows_this_session": 3361,
6
+ "trained_rows_total": 18961,
7
+ "total_db_rows": 238664,
8
+ "percentage": 7.9446418395736265,
9
  "final_loss": 0,
10
  "epochs": 3,
11
  "learning_rate": 5e-05,