Just some GGUF v2 quantizations of the model TinyLlama/tinyLlama-intermediate-checkpoints Step 720K pretrained on 1510B(1.5T) of tokens.