| | --- |
| | language: |
| | - en |
| | --- |
| | |
| | # SmolLM3 Training Configs |
| |
|
| | **[IMPORTANT NOTE]**: for the latest configs go to this repo: https://github.com/huggingface/smollm/tree/main/text/pretraining/smollm3 |
| |
|
| |
|
| | Here you can find the training configs for [SmoLLM3-3B-Base](https://huggingface.co/HuggingFaceTB/SmolLM3-3B-Base) using [nanotron](https://github.com/huggingface/nanotron/) with exact training details and data mixtures. |
| |
|
| | The model was trained on 11.2T tokens in 3 stages on 4k context: |
| |
|
| | - stage 1 [config](https://huggingface.co/datasets/HuggingFaceTB/smollm3-configs/blob/main/stage1_8T.yaml) |
| | - stage 2 [config](https://huggingface.co/datasets/HuggingFaceTB/smollm3-configs/blob/main/stage2_8T_9T.yaml) |
| | - stage 3 [config](https://huggingface.co/datasets/HuggingFaceTB/smollm3-configs/blob/main/stage3_9T_11T.yaml) |
| |
|
| |  |
| |
|
| |
|
| | And then we trained on an additional 2 stages to extend the contetx length to 64k: |
| |
|
| | - stage 4 [config](https://huggingface.co/datasets/HuggingFaceTB/smollm3-configs/blob/main/long_context_4k_to_32k.yaml) |
| | - stage 5 [config](https://huggingface.co/datasets/HuggingFaceTB/smollm3-configs/blob/main/long_context_32k_to_64.yaml) |
| |
|
| |  |
| |
|
| |
|
