trillionlabs
/

NemoSlides

Model card Files Files and versions

WonsukYangTL commited on Apr 21

Commit

d61905d

·

verified ·

1 Parent(s): a9ebc52

Update README.md

Files changed (1) hide show

README.md +3 -2

README.md CHANGED Viewed

@@ -17,7 +17,7 @@ base_model:
 **Nemotron Slide** is purpose-built to generate high-quality, aesthetic slides from a single instruction.
 ## Model Summary
 | Property | Value |
@@ -29,6 +29,7 @@ base_model:
 | **Precision** | bf16 |
 | **License** | Apache 2.0 |
 ## Post-Training Pipeline
 Starting from [NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16):
@@ -36,5 +37,5 @@ Starting from [NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://huggingface.co/nvidi
 1. **Supervised Fine-Tuning (SFT)**: We generated synthetic data via [NVIDIA Data Designer](https://build.nvidia.com/nemo/data-designer) and trained the model using [NVIDIA AutoModel](https://github.com/NVIDIA-NeMo/Automodel).
 2. **Direct Preference-Optimization**: We manually annotated slide generation results from diverse models, and further trained the model by [NVIDIA RL](https://github.com/NVIDIA-NeMo/RL).
 ## Evaluation Results

 **Nemotron Slide** is purpose-built to generate high-quality, aesthetic slides from a single instruction.
+---
 ## Model Summary
 | Property | Value |
 | **Precision** | bf16 |
 | **License** | Apache 2.0 |
+---
 ## Post-Training Pipeline
 Starting from [NVIDIA-Nemotron-3-Nano-30B-A3B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16):
 1. **Supervised Fine-Tuning (SFT)**: We generated synthetic data via [NVIDIA Data Designer](https://build.nvidia.com/nemo/data-designer) and trained the model using [NVIDIA AutoModel](https://github.com/NVIDIA-NeMo/Automodel).
 2. **Direct Preference-Optimization**: We manually annotated slide generation results from diverse models, and further trained the model by [NVIDIA RL](https://github.com/NVIDIA-NeMo/RL).
+---
 ## Evaluation Results