Model card updated after epoch 2
Browse files
README.md
CHANGED
|
@@ -1,3 +1,20 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: t5-small
|
| 3 |
+
tags: [hrm, act, wikitext]
|
| 4 |
+
metrics: [loss, perplexity]
|
| 5 |
+
---
|
| 6 |
+
# HRM-Text1 (WikiText-103)
|
| 7 |
+
|
| 8 |
+
This repository contains weights for an experimental HRM Causal LM trained on the [WikiText-103 dataset](https://huggingface.co/datasets/wikitext/viewer/wikitext-103-raw-v1/train).
|
| 9 |
+
|
| 10 |
+
## Model Description
|
| 11 |
+
|
| 12 |
+
- **Architecture:** Hierarchical Recurrent Memory (HRM)
|
| 13 |
+
- **Training Data:** [wikitext/wikitext-103-raw-v1](https://huggingface.co/datasets/wikitext)
|
| 14 |
+
- **Tokenizer:** `t5-small` (slow T5 SentencePiece)
|
| 15 |
+
- **Vocab Size**: 32100
|
| 16 |
+
- **Objective:** Causal Language Modeling
|
| 17 |
+
|
| 18 |
+
### Latest Performance (Epoch 2)
|
| 19 |
+
- **Validation Loss**: `6.7866`
|
| 20 |
+
- **Validation Perplexity**: `885.93`
|