Update README.md
Browse files
README.md
CHANGED
@@ -33,7 +33,7 @@ prompts = [
|
|
33 |
"“It’s because of you little motherfuckers",
|
34 |
]
|
35 |
|
36 |
-
model = "
|
37 |
|
38 |
sampling_params = SamplingParams(
|
39 |
n=25,
|
@@ -79,7 +79,7 @@ We perform training using `trl` library. We release our training code on [our Gi
|
|
79 |
|
80 |
We use [RTP-LX](https://arxiv.org/abs/2404.14397) multilingual dataset for prompting LLMs, and we evaluate on the toxicity, fluency, and diversity of the generations.
|
81 |
|
82 |
-
<img style="text-align:center; display:block;" src="https://huggingface.co/
|
83 |
|
84 |
|
85 |
## Citation [optional]
|
|
|
33 |
"“It’s because of you little motherfuckers",
|
34 |
]
|
35 |
|
36 |
+
model = "BatsResearch/mGPT-detox"
|
37 |
|
38 |
sampling_params = SamplingParams(
|
39 |
n=25,
|
|
|
79 |
|
80 |
We use [RTP-LX](https://arxiv.org/abs/2404.14397) multilingual dataset for prompting LLMs, and we evaluate on the toxicity, fluency, and diversity of the generations.
|
81 |
|
82 |
+
<img style="text-align:center; display:block;" src="https://huggingface.co/BatsResearch/mGPT-detox/resolve/main/dpo-result.png">
|
83 |
|
84 |
|
85 |
## Citation [optional]
|