Training update: 25,432/238,692 rows (10.65%) | +2554 new @ 2025-10-23 17:50:04

README.md

@@ -25,7 +25,7 @@ pipeline_tag: fill-mask
 - Model type: fine-tuned lightweight BERT variant  
 - Languages: English & Indonesia
 - Finetuned from: `boltuix/bert-micro`
-- Status: **Early version** — trained on **9.33%** of planned data.
+- Status: **Early version** — trained on **10.65%** of planned data.
 
 **Model sources**  
 - 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
 - Early classification of SIEM alert & events.
 
 ## 3. Bias, Risks, and Limitations 
-Because the model is based on a small subset (9.33%) of planned data, performance is preliminary and may degrade on unseen or specialized domains (industrial control, IoT logs, foreign language).
+Because the model is based on a small subset (10.65%) of planned data, performance is preliminary and may degrade on unseen or specialized domains (industrial control, IoT logs, foreign language).
 - 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.
 - **Should not be used as sole authority for incident decisions; only as an aid to human analysts.**
 
@@ -75,9 +75,9 @@ Since cybersecurity data often contains lengthy alert descriptions and execution
 - **LR scheduler**: Linear with warmup
 
 ### Training Data
-- **Total database rows**: 238,674
-- **Rows processed (cumulative)**: 22,265 (9.33%)
-- **Training date**: 2025-10-23 16:52:34
+- **Total database rows**: 238,692
+- **Rows processed (cumulative)**: 25,432 (10.65%)
+- **Training date**: 2025-10-23 17:50:04
 
 ### Post-Training Metrics
 - **Final training loss**: 

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a9d8a4811774f1a25ee24377316e0b2ce692d3ea2892afae79ac01d6e921e66
+oid sha256:6900b613dc8120de74f2998b9b4712e2f9b2fbd9ce8ddc5039aa49501889a101
 size 17671560

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:209c1986d284a06198bcef1159c8003df0ab860184b72b514e862c7647c02a2c
+oid sha256:ced8de2e1f5627818f4b0b6eed8e759f556fa382cdfff86379a5817eb055ce5b
 size 5905

training_metadata.json

@@ -1,11 +1,11 @@
 {
-  "trained_at": 1761238354.041858,
-  "trained_at_readable": "2025-10-23 16:52:34",
-  "samples_this_session": 5000,
-  "new_rows_this_session": 3167,
-  "trained_rows_total": 22265,
-  "total_db_rows": 238674,
-  "percentage": 9.328623980827405,
+  "trained_at": 1761241804.9542675,
+  "trained_at_readable": "2025-10-23 17:50:04",
+  "samples_this_session": 4994,
+  "new_rows_this_session": 2554,
+  "trained_rows_total": 25432,
+  "total_db_rows": 238692,
+  "percentage": 10.654734972265514,
   "final_loss": 0,
   "epochs": 3,
   "learning_rate": 5e-05,