7B Chinese Chatbot trained based on LLama2-base 7B (Pure LoRA Training)
Introduction
在完成了
Llama2-chat 7B Chinese 和
Llama2-chat 13B Chinese 的训练后，我非常好奇能否直接基于Llama2-base 系列直接进行SFT训练。这也是本模型仓库的初衷。
训练数据使用
BELLE项目中采样的50万SFT数据进行SFT训练。
After finishing the training of
Llama2-chat 7B Chinese and
Llama2-chat 13B Chinese, I am deeply intrigued by the possibility of conducting SFT (Style-Fine-Tuning) training directly based on the Llama2-base series. This is the fundamental purpose of this model repository.
Finally, after
RicardoLee/Llama2-base-7B-Chinese-50W-pre_release，
RicardoLee/Llama2-base-7B-Chinese-50W-Full2LoRA, I did find the right hyperparams to do the LoRA training stabelly based on Llama2-base 7B model. For more details please refer to the Train Detail section.
The training data is sampled from
BELLE project, which consists of 500,000 SFT samples.
Train Detail
一些训练上的细节：
- 训练框架：该模型使用了修改过的Chinese-LLaMA-Alpaca项目进行训练。
- Tokenizer：该模型使用了Chinese-Alpaca-Plus模型的tokenizer.model。这是因为LLama2本身的tokenizer.model同LLama1是一摸一样的。因此理论上可以完全复用Chinese-LLaMa项目的tokenizer而不会产生如何错位问题。
- 训练参数：该模型训练使用的超参数为：LoRA rank: 64, LR: 4e-4, Warmup ratio: 0.001.
- 训练资源：8卡V100。21小时
- 训练起始的loss：9.1402
- 训练终止的loss：1.4104
Some details in training:
- Trianing Framework: This model is trained on modified Chinese-LLaMA-Alpaca Framework.
- Tokenizer: This model utilizes the tokenizer.model from the Chinese-Alpaca-Plus model. The reason for this choice is that the tokenizer.model in LLama2 is identical to the one used in LLama1. As a result, it is theoretically feasible to entirely reuse the tokenizer from the Chinese-LLaMa project without encountering any issues related to token misalignment.
- Training Parameters: The hyperparams are: LoRA rank: 64, LR: 4e-4, Warmup ratio: 0.001.
- Training Resource: 8*V100, 21 hours.
- Initial Loss: 9.1402
- Train Loss: 1.4104
Inference
该模型依然采用stanford alpaca 模版。因此在测试时且别忘记添加开场白。开场白如下：
"Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n\n${Your Content}\n\n### Response:\n\n"
对于带上文的对话，开场白如下：
"Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n\nHuman:${Previous Human Content}\nAssistant:${Previous Assistance Content}\nHuman:${Your Question}\n\n### Response:\n\n"
This model still using the Stanford Alpaca template. Therefore, don't forget to add prologue template. The prologue template is:
"Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n\n${Your Content}\n\n### Response:\n\n"
For dialogue with context, the prelogue template is:
"Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n\nHuman:${Previous Human Content}\nAssistant:${Previous Machine Content}\nHuman:${Your Question}\n\n### Response:\n\n"
Licence
本仓库的模型依照 Apache-2.0 协议开源，模型的权重的使用则需要遵循LLama2
MODEL LICENCE。
This repository's models are open-sourced under the Apache-2.0 license, and their weight usage must adhere to LLama2
MODEL LICENCE license.
Future Work
将会在近期逐步放出
- 更大SFT数据规模训练下的模型。
- 13B及以下的LLama2 同LLama2-chat的模型，以供大家对比。
I will release the following models:
- Models trained on larger data scale.
- Models trained on LLama2 and LLama2-chat (under the 13B, since I only have V100), for comparison.
