Update README.md
Browse files
README.md
CHANGED
|
@@ -13,14 +13,14 @@ tags:
|
|
| 13 |
pipeline_tag: text-generation
|
| 14 |
---
|
| 15 |
|
| 16 |
-
# Atomight-V2.5-1.7B
|
| 17 |
|
| 18 |
<p align="center">
|
| 19 |
<img src="OfficialAtomight.png" alt="Atomight Logo" width="500" style="max-width: 100%;">
|
| 20 |
</p>
|
| 21 |
|
| 22 |
**Atomight V2.5 1.7B**, our *most capable and intelligent model* yet, is a multi-domain reasoning model built on **Qwen3-1.7B**, trained with **GRPO** (Group Relative Policy Optimization) across STEM, science, math, and code, using a curated six-dataset stack in a streaming round-robin setup — trained entirely on a single Colab free-tier T4 GPU.
|
| 23 |
-
|
| 24 |
## Overview
|
| 25 |
|
| 26 |
Atomight-V2.5 is post-trained from `Qwen/Qwen3-1.7B` (already instruct/chat-capable) using memory-efficient chunked GRPO to fit T4 VRAM constraints. Training data spans four domains, each selected for having genuinely verifiable rewards (exact-match answers or executable test cases) rather than relying purely on LLM-judged quality:
|
|
|
|
| 13 |
pipeline_tag: text-generation
|
| 14 |
---
|
| 15 |
|
| 16 |
+
# Atomight-V2.5-1.7B-C1
|
| 17 |
|
| 18 |
<p align="center">
|
| 19 |
<img src="OfficialAtomight.png" alt="Atomight Logo" width="500" style="max-width: 100%;">
|
| 20 |
</p>
|
| 21 |
|
| 22 |
**Atomight V2.5 1.7B**, our *most capable and intelligent model* yet, is a multi-domain reasoning model built on **Qwen3-1.7B**, trained with **GRPO** (Group Relative Policy Optimization) across STEM, science, math, and code, using a curated six-dataset stack in a streaming round-robin setup — trained entirely on a single Colab free-tier T4 GPU.
|
| 23 |
+
**Note: This is only the first checkpoint model of Atomight-V2.5-1.7B, but still usable and powerful model.**
|
| 24 |
## Overview
|
| 25 |
|
| 26 |
Atomight-V2.5 is post-trained from `Qwen/Qwen3-1.7B` (already instruct/chat-capable) using memory-efficient chunked GRPO to fit T4 VRAM constraints. Training data spans four domains, each selected for having genuinely verifiable rewards (exact-match answers or executable test cases) rather than relying purely on LLM-judged quality:
|