From e503b0404badee2c395030019ed48c27de51f7ae Mon Sep 17 00:00:00 2001 From: grad0s <84247753+grad0s@users.noreply.github.com> Date: Sat, 21 Dec 2024 07:54:10 +0200 Subject: [PATCH] Update smollm.md --- smollm.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/smollm.md b/smollm.md index e8ad5ced20..0700576547 100644 --- a/smollm.md +++ b/smollm.md @@ -136,7 +136,7 @@ The plot below compares Python-Edu to the unfiltered Python code and to using a SmolLM models are available in three sizes and were trained on the data mixture below: - 135M and 360M models, each trained on 600B tokens from [Smollm-Corpus](https://huggingface.co/datasets/HuggingFaceTB/smollm-corpus) -- 1.7B model, trained on 1T tokens from Smollm-Corpus +- 1.7B model, trained on 11T tokens from Smollm-Corpus