Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: mit
|
| 5 |
+
tags:
|
| 6 |
+
- pytorch
|
| 7 |
+
- causal-lm
|
| 8 |
+
- checkpoints
|
| 9 |
+
- sft
|
| 10 |
+
- maria-1b
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Maria-1B SFT & Alignment Checkpoints
|
| 14 |
+
|
| 15 |
+
This repository hosts all authentic PyTorch `.pt` checkpoints trained during the SFT and alignment phase of **Maria-1B** (HyperGQE architecture, 1.07B parameters).
|
| 16 |
+
|
| 17 |
+
## Available Checkpoints
|
| 18 |
+
|
| 19 |
+
| Filename | Format | Description |
|
| 20 |
+
| :--- | :--- | :--- |
|
| 21 |
+
| `maria_1b_unified_sft_v6.pt` | PyTorch State Dict | Final Unified SFT v6 checkpoint (Analyst + Dialogue + Coding) |
|
| 22 |
+
| `maria_1b_unified_sft_v5.pt` | PyTorch State Dict | Unified SFT v5 checkpoint |
|
| 23 |
+
| `maria_1b_20B_analyst_v4.pt` | PyTorch State Dict | Domain reasoning & analytics specialist v4 |
|
| 24 |
+
| `maria_1b_20B_analyst_v3.pt` | PyTorch State Dict | Domain reasoning & analytics specialist v3 |
|
| 25 |
+
| `maria_1b_20B_instruct_v2.pt` | PyTorch State Dict | Conversational Chat & Instruction tuning v2 |
|
| 26 |
+
| `maria_1b_20B_instruct_chat.pt`| PyTorch State Dict | Initial conversational patch checkpoint |
|
| 27 |
+
|
| 28 |
+
## Architecture & Base Model
|
| 29 |
+
* **Base Model**: [Zahid2005/Maria-1B](https://huggingface.co/Zahid2005/Maria-1B) (20B token scratch pretraining)
|
| 30 |
+
* **Instruct Model**: [Zahid2005/Maria-1B-Instruct](https://huggingface.co/Zahid2005/Maria-1B-Instruct)
|
| 31 |
+
* **Preprint Citation**: [https://doi.org/10.5281/zenodo.22747827](https://doi.org/10.5281/zenodo.22747827)
|