Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,39 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
language:
|
| 5 |
+
- en
|
| 6 |
+
tags:
|
| 7 |
+
- causal-lm
|
| 8 |
+
- text-generation
|
| 9 |
+
- pytorch
|
| 10 |
+
- safetensors
|
| 11 |
+
- Lumen
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# 🧠 LumenBase
|
| 15 |
+
|
| 16 |
+
**LumenBase** is a 128M parameter language model developed from scratch by Hariom Jangra.
|
| 17 |
+
It serves as the **base model** for the *Lumen* series, designed for research and learning Purposes.
|
| 18 |
+
|
| 19 |
+
---
|
| 20 |
+
|
| 21 |
+
## 🧩 Model Details
|
| 22 |
+
|
| 23 |
+
- **Architecture:** Decoder-only Transformer
|
| 24 |
+
- **Parameters:** 128 million
|
| 25 |
+
- **Precision:** Float32 (safetensors format)
|
| 26 |
+
- **Framework:** PyTorch
|
| 27 |
+
- **File:** `LumenBase.safetensors`
|
| 28 |
+
- **Tokenizer:** BPE-based
|
| 29 |
+
- **Trained from scratch:** Yes
|
| 30 |
+
- **Post-trained version:** Coming soon
|
| 31 |
+
|
| 32 |
+
---
|
| 33 |
+
|
| 34 |
+
## 🧠 Intended Use
|
| 35 |
+
|
| 36 |
+
- Research & experimentation
|
| 37 |
+
- Educational purposes (understanding Transformer architectures)
|
| 38 |
+
|
| 39 |
+
|