Update README.md
Browse files
README.md
CHANGED
|
@@ -12,6 +12,7 @@ tags:
|
|
| 12 |
- rlhf
|
| 13 |
datasets:
|
| 14 |
- Anthropic/hh-rlhf
|
|
|
|
| 15 |
---
|
| 16 |
|
| 17 |
# NV-Llama2-70B-RLHF-Chat
|
|
@@ -183,4 +184,4 @@ Pre-requisite: You would need at least a machine with 4 40GB or 2 80GB NVIDIA GP
|
|
| 183 |
## Limitations
|
| 184 |
- The model was trained on the data that contains toxic language and societal biases originally crawled from the Internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts.
|
| 185 |
- The Model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.
|
| 186 |
-
- We recommend deploying the model with [NeMo Guardrails](https://github.com/NVIDIA/NeMo-Guardrails) to mitigate these potential issues.
|
|
|
|
| 12 |
- rlhf
|
| 13 |
datasets:
|
| 14 |
- Anthropic/hh-rlhf
|
| 15 |
+
- nvidia/sft_datablend_v1
|
| 16 |
---
|
| 17 |
|
| 18 |
# NV-Llama2-70B-RLHF-Chat
|
|
|
|
| 184 |
## Limitations
|
| 185 |
- The model was trained on the data that contains toxic language and societal biases originally crawled from the Internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts.
|
| 186 |
- The Model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.
|
| 187 |
+
- We recommend deploying the model with [NeMo Guardrails](https://github.com/NVIDIA/NeMo-Guardrails) to mitigate these potential issues.
|