final SFT PRO MIX from best pretrain checkpoint

training-runs/sft-italian-clean-1b/checkpoint_last.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6afe655128b187b85e76322a3399c7f2863b727bd8a29a380f72302f4a30e9bf
+oid sha256:2e35c45fea3720aca3341bd0c48321bf79f45dfc56dac1b3e7e73b525b1373ca
 size 15758460499

training-runs/sft-italian-clean-1b/tokenizer/tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": "<|endoftext|>",
+  "eos_token": "<|endoftext|>",
+  "errors": "replace",
+  "is_local": false,
+  "local_files_only": false,
+  "model_max_length": 1024,
+  "pad_token": "<|endoftext|>",
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>"
+}