Create README.md

README.md

@@ -0,0 +1,33 @@
+---
+language:
+- zh
+- en
+- ja
+---
+Dataset for fine-tuning the LLM Quirel.
+
+The code for processing data will be open-sourced later.
+
+# [cn_deepctrl_sft]
+
+Collected from https://www.modelscope.cn/datasets/deepctrl/deepctrl-sft-data
+
+# [en_deepctrl_sft]
+
+Collected from https://www.modelscope.cn/datasets/deepctrl/deepctrl-sft-data
+
+# [ja_databricks_dolly_15k]
+
+Collected from https://huggingface.co/datasets/kunishou/databricks-dolly-15k-ja
+
+# [ja_hh_rlhf_49k]
+
+Collected from https://huggingface.co/datasets/fujiki/japanese_hh-rlhf-49k
+
+# [ja_ichikara_instruction]
+
+Collected from https://huggingface.co/datasets/msfm/ichikara-instruction-all
+
+# [ja_oasst2_chat_68k]
+
+Collected from https://huggingface.co/datasets/kunishou/oasst2-chat-68k-ja