update card
Browse files
README.md
CHANGED
|
@@ -11,8 +11,6 @@ tags:
|
|
| 11 |
|
| 12 |
Reward model trained for PPO for VK NLP course.
|
| 13 |
|
| 14 |
-
Is likely to be overfitted.
|
| 15 |
-
|
| 16 |
## Model Details
|
| 17 |
|
| 18 |
### Model Description
|
|
@@ -20,6 +18,8 @@ Is likely to be overfitted.
|
|
| 20 |
<!-- Provide a longer summary of what this model is. -->
|
| 21 |
The model is LLM [HuggingFaceTB/SmolLM-135M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM-135M-Instruct).
|
| 22 |
|
|
|
|
|
|
|
| 23 |
Model trained with TRL for PPO trainig.
|
| 24 |
|
| 25 |
### Model Sources
|
|
|
|
| 11 |
|
| 12 |
Reward model trained for PPO for VK NLP course.
|
| 13 |
|
|
|
|
|
|
|
| 14 |
## Model Details
|
| 15 |
|
| 16 |
### Model Description
|
|
|
|
| 18 |
<!-- Provide a longer summary of what this model is. -->
|
| 19 |
The model is LLM [HuggingFaceTB/SmolLM-135M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM-135M-Instruct).
|
| 20 |
|
| 21 |
+
Trained only last linear layer.
|
| 22 |
+
|
| 23 |
Model trained with TRL for PPO trainig.
|
| 24 |
|
| 25 |
### Model Sources
|