Training update: 2,116/237,881 rows (0.89%) | +500 new @ 2025-10-21 06:00:02

README.md

@@ -21,7 +21,7 @@ library_name: transformers
 - Model type: fine-tuned lightweight BERT variant  
 - Languages: English & Indonesia
 - Finetuned from: `boltuix/bert-micro`
-- Status: **Early version** — trained on **0.68%** of planned data.
+- Status: **Early version** — trained on **0.89%** of planned data.
 
 **Model sources**  
 - Base model: [boltuix/bert-micro](https://huggingface.co/boltuix/bert-micro)
@@ -42,7 +42,7 @@ You can use this model to classify cybersecurity-related text — for example, w
 - Not tested for non-cybersecurity domains or out-of-distribution data.
 
 ## 3. Bias, Risks, and Limitations 
-Because the model is based on a small subset (0.68%) 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 (0.89%) 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.
@@ -62,8 +62,8 @@ predicted_class = logits.argmax(dim=-1).item()
 ```
 
 ## 5. Training Details
-- **Trained records**: 1,616 / 237,718 (0.68%)
+- **Trained records**: 2,116 / 237,881 (0.89%)
 - **Learning rate**: 5e-05
 - **Epochs**: 3
-- **Batch size**: 8
+- **Batch size**: 16
 - **Max sequence length**: 512

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:77bacaf8a88c4d1a2ec9600b44d7c70b6f0a0779ba39204c66e7560f2822fa01
+oid sha256:a7ce222f46cef58ed19978bf1a4b88f5447d4d114d7ad81f61ed702f8f3c9d9c
 size 17671560

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:85b878d7ad871ddbb6534d459d90aa4c48bd16f59636bcec6c7495740a5d8ccf
+oid sha256:930fff7680df5b08cacf7b90bbbc38051edff1aea340c88857c47f5238ec8dc4
 size 5905

training_metadata.json

@@ -1,11 +1,11 @@
 {
-  "trained_at": 1761021667.0032144,
-  "trained_at_readable": "2025-10-21 04:41:07",
-  "samples_this_session": 4634,
+  "trained_at": 1761026402.3035963,
+  "trained_at_readable": "2025-10-21 06:00:02",
+  "samples_this_session": 960,
   "new_rows_this_session": 500,
-  "trained_rows_total": 1616,
-  "total_db_rows": 237718,
-  "percentage": 0.679797070478466,
+  "trained_rows_total": 2116,
+  "total_db_rows": 237881,
+  "percentage": 0.8895203904473246,
   "final_loss": 0,
   "epochs": 3,
   "learning_rate": 5e-05