Update README.md
Browse files
README.md
CHANGED
|
@@ -53,10 +53,9 @@ Introduction to Finetuning LLMs course - Learning
|
|
| 53 |
### Downstream Use [optional]
|
| 54 |
|
| 55 |
This model is designed for tasks requiring improved alignment with human preferences, such as:
|
| 56 |
-
Chatbots
|
| 57 |
-
Question-answering systems
|
| 58 |
-
General text generation with enhanced preference alignment<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 59 |
-
|
| 60 |
|
| 61 |
### Out-of-Scope Use
|
| 62 |
|
|
@@ -166,12 +165,6 @@ Training Data
|
|
| 166 |
- Content: Prompts, chosen answers, and rejected answers
|
| 167 |
|
| 168 |
|
| 169 |
-
|
| 170 |
-
| Tasks |Version|Filter|n-shot| Metric | |Value | |Stderr|
|
| 171 |
-
|---------|------:|------|-----:|--------|---|-----:|---|-----:|
|
| 172 |
-
|hellaswag| 1|none | 0|acc |↑ |0.4516|± |0.0050|
|
| 173 |
-
| | |none | 0|acc_norm|↑ |0.6139|± |0.0049|
|
| 174 |
-
|
| 175 |
Task: HellaSwag
|
| 176 |
- This is a benchmark task designed to evaluate a model's commonsense reasoning and ability to complete scenarios logically.
|
| 177 |
- No specific filtering was applied to the test set.
|
|
@@ -190,8 +183,6 @@ Interpretation:
|
|
| 190 |
|
| 191 |
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 192 |
|
| 193 |
-
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 194 |
-
|
| 195 |
- **Hardware Type:** A100
|
| 196 |
- **Hours used:** No comment
|
| 197 |
- **Cloud Provider:** Google Collab
|
|
@@ -203,7 +194,7 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
| 203 |
Hardware: A100 GPU
|
| 204 |
|
| 205 |
|
| 206 |
-
## Model Card
|
| 207 |
|
| 208 |
Ruth Shacterman
|
| 209 |
|
|
|
|
| 53 |
### Downstream Use [optional]
|
| 54 |
|
| 55 |
This model is designed for tasks requiring improved alignment with human preferences, such as:
|
| 56 |
+
- Chatbots
|
| 57 |
+
- Question-answering systems
|
| 58 |
+
- General text generation with enhanced preference alignment<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
|
|
|
| 59 |
|
| 60 |
### Out-of-Scope Use
|
| 61 |
|
|
|
|
| 165 |
- Content: Prompts, chosen answers, and rejected answers
|
| 166 |
|
| 167 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 168 |
Task: HellaSwag
|
| 169 |
- This is a benchmark task designed to evaluate a model's commonsense reasoning and ability to complete scenarios logically.
|
| 170 |
- No specific filtering was applied to the test set.
|
|
|
|
| 183 |
|
| 184 |
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 185 |
|
|
|
|
|
|
|
| 186 |
- **Hardware Type:** A100
|
| 187 |
- **Hours used:** No comment
|
| 188 |
- **Cloud Provider:** Google Collab
|
|
|
|
| 194 |
Hardware: A100 GPU
|
| 195 |
|
| 196 |
|
| 197 |
+
## Model Card Author
|
| 198 |
|
| 199 |
Ruth Shacterman
|
| 200 |
|