Update README.md
Browse files
README.md
CHANGED
|
@@ -2,7 +2,7 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
|
| 5 |
-
#
|
| 6 |
|
| 7 |
**[autonomousX/instinct-1.4B on Hugging Face](https://huggingface.co/autonomousX)**
|
| 8 |
|
|
@@ -122,7 +122,7 @@ The loss curves are saved in `training_log.txt` and `val_perplexity.txt`. Below
|
|
| 122 |
|
| 123 |
The entire pipeline used to train the model is fully reproducible. This includes the dataset pipeline, tokenizer creation, model architecture, TPU training loop, and checkpointing system.
|
| 124 |
|
| 125 |
-
**Full training pipeline repository:** [train.py](https://github.com/YADAV1825/
|
| 126 |
|
| 127 |
---
|
| 128 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
|
| 5 |
+
# Instinct 1.4B
|
| 6 |
|
| 7 |
**[autonomousX/instinct-1.4B on Hugging Face](https://huggingface.co/autonomousX)**
|
| 8 |
|
|
|
|
| 122 |
|
| 123 |
The entire pipeline used to train the model is fully reproducible. This includes the dataset pipeline, tokenizer creation, model architecture, TPU training loop, and checkpointing system.
|
| 124 |
|
| 125 |
+
**Full training pipeline repository:** [train.py](https://github.com/YADAV1825/Instinct 1.4B/blob/main/train.py) (Example link format)
|
| 126 |
|
| 127 |
---
|
| 128 |
|