Update README.md
Browse files
README.md
CHANGED
|
@@ -32,7 +32,7 @@ Nemotron-Math-Proofs-v3-SFT is a long-form mathematical reasoning dataset contai
|
|
| 32 |
|
| 33 |
The source pool contains 15,879 hard proof problems selected from the [AoPS](https://artofproblemsolving.com/) subset of [nvidia/Nemotron-Math-Proofs-v1](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v1). Responses are generated using [DeepSeek-V4-Pro](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro) in Max inference mode. A four-round generate-verify-refine pipeline produces initial proofs in round 1 and refinement trajectories for problems that remain unsolved in rounds 2 through 4. The pipeline also produces verifier and meta-verifier traces following the proof-generation and proof-verification prompting style described in the [DeepSeekMath-V2 paper](https://arxiv.org/abs/2511.22570).
|
| 34 |
|
| 35 |
-
Full details about dataset construction can be found in our technical report [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://
|
| 36 |
|
| 37 |
This dataset is ready for commercial or non-commercial uses.
|
| 38 |
|
|
@@ -141,7 +141,7 @@ Total Recorded Token Count: 39,047,191,715
|
|
| 141 |
Generated token counts are taken from the inference metadata. Input token counts are pipeline estimates based on prompt text length rather than a tokenizer re-count.
|
| 142 |
|
| 143 |
## Reference(s):
|
| 144 |
-
* [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://
|
| 145 |
* [DeepSeek-V4-Pro](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro)
|
| 146 |
* [nvidia/Nemotron-Math-Proofs-v1](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v1)
|
| 147 |
* [nvidia/Nemotron-Math-Proofs-v2](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v2)
|
|
|
|
| 32 |
|
| 33 |
The source pool contains 15,879 hard proof problems selected from the [AoPS](https://artofproblemsolving.com/) subset of [nvidia/Nemotron-Math-Proofs-v1](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v1). Responses are generated using [DeepSeek-V4-Pro](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro) in Max inference mode. A four-round generate-verify-refine pipeline produces initial proofs in round 1 and refinement trajectories for problems that remain unsolved in rounds 2 through 4. The pipeline also produces verifier and meta-verifier traces following the proof-generation and proof-verification prompting style described in the [DeepSeekMath-V2 paper](https://arxiv.org/abs/2511.22570).
|
| 34 |
|
| 35 |
+
Full details about dataset construction can be found in our technical report [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://arxiv.org/abs/2609.10712).
|
| 36 |
|
| 37 |
This dataset is ready for commercial or non-commercial uses.
|
| 38 |
|
|
|
|
| 141 |
Generated token counts are taken from the inference metadata. Input token counts are pipeline estimates based on prompt text length rather than a tokenizer re-count.
|
| 142 |
|
| 143 |
## Reference(s):
|
| 144 |
+
* [An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics](https://arxiv.org/abs/2609.10712) (technical report)
|
| 145 |
* [DeepSeek-V4-Pro](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro)
|
| 146 |
* [nvidia/Nemotron-Math-Proofs-v1](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v1)
|
| 147 |
* [nvidia/Nemotron-Math-Proofs-v2](https://huggingface.co/datasets/nvidia/Nemotron-Math-Proofs-v2)
|