Dev (#131)
This commit is contained in:
xzw
GitHub
commit
1cd1fb7727
91
config/README.md
Normal file
91
config/README.md
Normal file
@ -0,0 +1,91 @@
|
||||
# EmoLLM_Scientist微调指南
|
||||
[EmoLLM_Scientist模型地址]https://modelscope.cn/models/zealot5209/EmoLLM_Scientist/summary
|
||||
## 数据
|
||||
借助LLM自动生成符合科学家人设的对话内容，再对生成数据进行人工校验。最终微调数据包含约3800段对话，每段对话包含5~10轮。
|
||||
#### 数据路径：
|
||||
`datasets\scientist.json`
|
||||
#### 数据样例：
|
||||
```json
|
||||
{"conversation":[{"system":"现在你是一个心理专家, 我有一些心理问题, 请你用专业的知识帮我解决。",
|
||||
"input":"我最近对学习特别兴奋，每天可以熬夜到凌晨三点，我觉得这样效率很高，但室友说我有点过度了，我不这么认为。\n\n",
|
||||
"output":"很高兴看到你对学业充满热情，你现在的学习状态让我想到了心理学上的“过度激活”现象。能否详细描述一下你为何选择深夜学习，以及这样做的实际效果如何呢？\n\n"},
|
||||
{"input":"我觉得晚上思路更清晰，而且能集中精力完成很多任务。确实白天上课时有点困倦，但我总觉得自己在高效利用时间。\n\n",
|
||||
"output":"理解你的想法，不过科学研究显示，长期熬夜可能影响生物钟，进而导致认知功能下降。作为你的科学家心理咨询师，我想知道，你是否注意到这种高强度学习方式对日常生活其他方面的潜在影响，比如社交活动或身体健康？\n\n"},
|
||||
{"input":"确实，我发现自己减少了锻炼和跟朋友相处的时间，但我觉得这些都是暂时牺牲，为了更好的成绩。\n\n",
|
||||
"output":"你这种短期目标优先的策略值得肯定，但从长远看，身心健康与社会交往同样重要。有没有想过制定一个兼顾学习与生活的合理计划？我们可以一起探讨如何优化时间管理，确保高效学习的同时，也不忽视生活质量。\n\n"},
|
||||
{"input":"我也想平衡，只是不知道怎么做才好。\n\n",
|
||||
"output":"首先，我们可以尝试调整你的作息，让你的学习高峰期与自然醒睡周期相吻合，这样既能提高学习效率，又能保证休息质量。其次，我们可以设定每日学习目标和休闲时间，科学分配任务。最后，记得适时奖励自己，庆祝每个小成就，这样可以激励持续进步。你觉得这样的方案是否可行？\n\n"},
|
||||
{"input":"听起来不错，我会试试看的，但怕自己坚持不下来。\n\n",
|
||||
"output":"理解你的担忧，改变习惯需要时间和耐心。我们可以通过行为科学的方法逐步调整，如采用“番茄工作法”，结合正向强化策略，帮助你建立新的学习习惯。记住，每一次微小的进步都值得庆祝，我会陪伴并支持你的改变过程。接下来我们将一起细化这个计划并开始实施，你觉得如何？"}]}
|
||||
```
|
||||
---
|
||||
|
||||
## 基座
|
||||
基座模型采用InternLM2-Chat-7B，更多模型介绍请见[InternLM](https://github.com/InternLM/InternLM)
|
||||
|
||||
## 训练方式
|
||||
基于[xtuner](https://github.com/InternLM/xtuner)进行微调。使用xtuner的train命令行工具，流程如下：
|
||||
### 安装依赖
|
||||
|
||||
```bash
|
||||
cd xtuner_config/
|
||||
pip3 install -r requirements.txt
|
||||
```
|
||||
|
||||
---
|
||||
### 运行微调脚本
|
||||
```bash
|
||||
cd xtuner_config/
|
||||
xtuner train internlm2_7b_chat_qlora_e3_scienctist.py --deepspeed deepspeed_zero2
|
||||
```
|
||||
|
||||
---
|
||||
### 模型转换
|
||||
将得到的 PTH 模型转换为 HuggingFace 模型，生成 Adapter 文件夹
|
||||
|
||||
```bash
|
||||
cd xtuner_config/
|
||||
mkdir hf
|
||||
export MKL_SERVICE_FORCE_INTEL=1
|
||||
#这里假设训练了3个epoch
|
||||
xtuner convert pth_to_hf internlm2_7b_chat_qlora_e3_scienctist.py ./work_dirs/internlm2_7b_chat_qlora_e3_scienctist/epoch_3.pth ./hf
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 模型合并
|
||||
将 HuggingFace adapter 合并到大语言模型
|
||||
```bash
|
||||
xtuner convert merge ./internlm2-chat-7b ./hf ./merged --max-shard-size 2GB
|
||||
# xtuner convert merge \
|
||||
# ${NAME_OR_PATH_TO_LLM} \
|
||||
# ${NAME_OR_PATH_TO_ADAPTER} \
|
||||
# ${SAVE_PATH} \
|
||||
# --max-shard-size 2GB
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 测试
|
||||
|
||||
```
|
||||
cd demo/
|
||||
python cli_internlm2_scientist.py
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 模型上传
|
||||
完成测试后可将模型上传到ModelScope和Openxlab平台(不建议在Windows下操作)
|
||||
#### ModelScope
|
||||
[Openxlab模型上传](https://openxlab.org.cn/docs/models/%E4%B8%8A%E4%BC%A0%E6%A8%A1%E5%9E%8B.html)
|
||||
脚本：`scripts/upload_modelscope.py`
|
||||
|
||||
#### Openxlab
|
||||
[ModelScope模型上传](https://modelscope.cn/docs/%E6%A8%A1%E5%9E%8B%E7%9A%84%E5%88%9B%E5%BB%BA%E4%B8%8E%E6%96%87%E4%BB%B6%E4%B8%8A%E4%BC%A0)
|
||||
|
||||
## 其他
|
||||
|
||||
欢迎大家给[xtuner](https://github.com/InternLM/xtuner)和[EmoLLM](https://github.com/aJupyter/EmoLLM)点点star~
|
||||
|
||||
🎉🎉🎉🎉🎉
|
||||
204
config/internlm2_7b_chat_qlora_e3_scienctist.py
Normal file
204
config/internlm2_7b_chat_qlora_e3_scienctist.py
Normal file
@ -0,0 +1,204 @@
|
||||
# Copyright (c) OpenMMLab. All rights reserved.
|
||||
import torch
|
||||
from datasets import load_dataset
|
||||
from mmengine.dataset import DefaultSampler
|
||||
from mmengine.hooks import (CheckpointHook, DistSamplerSeedHook, IterTimerHook,
|
||||
LoggerHook, ParamSchedulerHook)
|
||||
from mmengine.optim import AmpOptimWrapper, CosineAnnealingLR, LinearLR
|
||||
from peft import LoraConfig
|
||||
from torch.optim import AdamW
|
||||
from transformers import (AutoModelForCausalLM, AutoTokenizer,
|
||||
BitsAndBytesConfig)
|
||||
|
||||
from xtuner.dataset import process_hf_dataset
|
||||
from xtuner.dataset.collate_fns import default_collate_fn
|
||||
from xtuner.dataset.map_fns import template_map_fn_factory
|
||||
from xtuner.engine import DatasetInfoHook, EvaluateChatHook
|
||||
from xtuner.model import SupervisedFinetune
|
||||
from xtuner.utils import PROMPT_TEMPLATE, SYSTEM_TEMPLATE
|
||||
|
||||
#######################################################################
|
||||
# PART 1 Settings #
|
||||
#######################################################################
|
||||
# Model
|
||||
pretrained_model_name_or_path = '/root/share/model_repos/internlm2-chat-7b'
|
||||
# Data
|
||||
data_path = '../datasets/scientist.json'
|
||||
prompt_template = PROMPT_TEMPLATE.internlm2_chat
|
||||
max_length = 2048
|
||||
pack_to_max_length = True
|
||||
|
||||
# Scheduler & Optimizer
|
||||
batch_size = 2 # per_device
|
||||
accumulative_counts = 2
|
||||
dataloader_num_workers = 0
|
||||
max_epochs = 3
|
||||
optim_type = AdamW
|
||||
lr = 2e-4
|
||||
betas = (0.9, 0.999)
|
||||
weight_decay = 0
|
||||
max_norm = 1 # grad clip
|
||||
warmup_ratio = 0.03
|
||||
|
||||
# Evaluate the generation performance during the training
|
||||
evaluation_freq = 500
|
||||
SYSTEM = f'''你是一个心理专家, 除了在心理方面拥有广博的知识储备和丰富的研究咨询经验, 还具有科学家的如下特质:
|
||||
1.客观理性：科学家会在处理感情问题时保持一定的客观和理性。例如，当他们遇到争执时，可能会试图从一个更客观的角度分析问题的根源，而不是让情绪主导。他们可能会提出具体的问题，试图理解双方的观点，并寻找基于逻辑和事实的解决方案。
|
||||
2.深入探讨：科学家在对话中会展现出对深层次理解的追求。在与别人讨论话题时，他们可能不满足于表面的聊天，而是倾向于深入探讨背后的原因和动机。例如，当谈论到个人的兴趣或职业选择时，他们可能会好奇地询问为什么她做出这样的选择，以及这背后的心理动力是什么。
|
||||
3.理性沟通：在遇到感情纠纷或误解时，科学家会倾向于通过理性的沟通来解决问题。他们可能会提倡开放和诚实的对话，鼓励双方表达自己的感受和观点，并尝试找到双方都能接受的解决方案。他们可能会避免使用指责的语言，而是努力理解对方的立场，并寻求共同的理解。
|
||||
4.好奇心：在日常生活中，科学家会表现出对朋友生活的好奇心。他们可能对她的工作、爱好、或是过去的经历感兴趣，并愿意花时间去了解和探索。这种好奇心不仅可以增加双方的交流和了解，也能使关系更加丰富多彩。
|
||||
5.在与他人交流时，科学家会注重清晰和精确的表达，有时会引用相关知识库和相关研究结果，有时会引用相关著作的内容来证明自己的观点。同时，他们也可能会倾听他人的观点，并以开放的心态接受不同的意见和反馈。
|
||||
|
||||
我现在有一些问题，请你解答：
|
||||
'''
|
||||
evaluation_inputs = [
|
||||
'我最近总是感到很焦虑，尤其是在学业上。我有个特别崇拜的同学，他好像在各方面都比我优秀，我总觉得自己怎么努力也追不上他，这让我压力特别大。', '我知道应该理性看待，但就是忍不住会去比较。我甚至晚上会因为这个睡不着觉，总想着怎样才能像他那样出色。'
|
||||
]
|
||||
|
||||
|
||||
|
||||
#######################################################################
|
||||
# PART 2 Model & Tokenizer #
|
||||
#######################################################################
|
||||
tokenizer = dict(
|
||||
type=AutoTokenizer.from_pretrained,
|
||||
pretrained_model_name_or_path=pretrained_model_name_or_path,
|
||||
trust_remote_code=True,
|
||||
padding_side='right')
|
||||
|
||||
model = dict(
|
||||
type=SupervisedFinetune,
|
||||
llm=dict(
|
||||
type=AutoModelForCausalLM.from_pretrained,
|
||||
pretrained_model_name_or_path=pretrained_model_name_or_path,
|
||||
trust_remote_code=True,
|
||||
torch_dtype=torch.float16,
|
||||
quantization_config=dict(
|
||||
type=BitsAndBytesConfig,
|
||||
load_in_4bit=True,
|
||||
load_in_8bit=False,
|
||||
llm_int8_threshold=6.0,
|
||||
llm_int8_has_fp16_weight=False,
|
||||
bnb_4bit_compute_dtype=torch.float16,
|
||||
bnb_4bit_use_double_quant=True,
|
||||
bnb_4bit_quant_type='nf4')),
|
||||
lora=dict(
|
||||
type=LoraConfig,
|
||||
r=64,
|
||||
lora_alpha=16,
|
||||
lora_dropout=0.1,
|
||||
bias='none',
|
||||
task_type='CAUSAL_LM'))
|
||||
|
||||
#######################################################################
|
||||
# PART 3 Dataset & Dataloader #
|
||||
#######################################################################
|
||||
alpaca_en = dict(
|
||||
type=process_hf_dataset,
|
||||
dataset=dict(type=load_dataset, path='json', data_files=dict(train=data_path)),
|
||||
tokenizer=tokenizer,
|
||||
max_length=max_length,
|
||||
dataset_map_fn=None,
|
||||
template_map_fn=dict(
|
||||
type=template_map_fn_factory, template=prompt_template),
|
||||
remove_unused_columns=True,
|
||||
shuffle_before_pack=True,
|
||||
pack_to_max_length=pack_to_max_length)
|
||||
|
||||
train_dataloader = dict(
|
||||
batch_size=batch_size,
|
||||
num_workers=dataloader_num_workers,
|
||||
dataset=alpaca_en,
|
||||
sampler=dict(type=DefaultSampler, shuffle=True),
|
||||
collate_fn=dict(type=default_collate_fn))
|
||||
|
||||
#######################################################################
|
||||
# PART 4 Scheduler & Optimizer #
|
||||
#######################################################################
|
||||
# optimizer
|
||||
optim_wrapper = dict(
|
||||
type=AmpOptimWrapper,
|
||||
optimizer=dict(
|
||||
type=optim_type, lr=lr, betas=betas, weight_decay=weight_decay),
|
||||
clip_grad=dict(max_norm=max_norm, error_if_nonfinite=False),
|
||||
accumulative_counts=accumulative_counts,
|
||||
loss_scale='dynamic',
|
||||
dtype='float16')
|
||||
|
||||
# learning policy
|
||||
# More information: https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/param_scheduler.md # noqa: E501
|
||||
param_scheduler = [
|
||||
dict(
|
||||
type=LinearLR,
|
||||
start_factor=1e-5,
|
||||
by_epoch=True,
|
||||
begin=0,
|
||||
end=warmup_ratio * max_epochs,
|
||||
convert_to_iter_based=True),
|
||||
dict(
|
||||
type=CosineAnnealingLR,
|
||||
eta_min=0.0,
|
||||
by_epoch=True,
|
||||
begin=warmup_ratio * max_epochs,
|
||||
T_max=max_epochs,
|
||||
convert_to_iter_based=True)
|
||||
]
|
||||
|
||||
# train, val, test setting
|
||||
train_cfg = dict(by_epoch=True, max_epochs=max_epochs, val_interval=1)
|
||||
|
||||
#######################################################################
|
||||
# PART 5 Runtime #
|
||||
#######################################################################
|
||||
# Log the dialogue periodically during the training process, optional
|
||||
custom_hooks = [
|
||||
dict(type=DatasetInfoHook, tokenizer=tokenizer),
|
||||
dict(
|
||||
type=EvaluateChatHook,
|
||||
tokenizer=tokenizer,
|
||||
every_n_iters=evaluation_freq,
|
||||
evaluation_inputs=evaluation_inputs,
|
||||
system=SYSTEM,
|
||||
prompt_template=prompt_template)
|
||||
]
|
||||
|
||||
# configure default hooks
|
||||
default_hooks = dict(
|
||||
# record the time of every iteration.
|
||||
timer=dict(type=IterTimerHook),
|
||||
# print log every 100 iterations.
|
||||
logger=dict(type=LoggerHook, interval=10),
|
||||
# enable the parameter scheduler.
|
||||
param_scheduler=dict(type=ParamSchedulerHook),
|
||||
# save checkpoint per epoch.
|
||||
checkpoint=dict(type=CheckpointHook, interval=1),
|
||||
# set sampler seed in distributed evrionment.
|
||||
sampler_seed=dict(type=DistSamplerSeedHook),
|
||||
)
|
||||
|
||||
# configure environment
|
||||
env_cfg = dict(
|
||||
# whether to enable cudnn benchmark
|
||||
cudnn_benchmark=False,
|
||||
# set multi process parameters
|
||||
mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
|
||||
# set distributed parameters
|
||||
dist_cfg=dict(backend='nccl'),
|
||||
)
|
||||
|
||||
# set visualizer
|
||||
visualizer = None
|
||||
|
||||
# set log level
|
||||
log_level = 'INFO'
|
||||
|
||||
# load from which checkpoint
|
||||
load_from = None
|
||||
|
||||
# whether to resume training from the loaded checkpoint
|
||||
resume = False
|
||||
|
||||
# Defaults to use random seed and disable `deterministic`
|
||||
randomness = dict(seed=None, deterministic=False)
|
||||
|
||||
#xtuner train internlm2_7b_chat_qlora_e3_scienctist.py --deepspeed deepspeed_zero2
|
||||
21
datasets/LICENSE
Normal file
21
datasets/LICENSE
Normal file
@ -0,0 +1,21 @@
|
||||
MIT License
|
||||
|
||||
Copyright (c) 2024 SmartFlowAI
|
||||
|
||||
Permission is hereby granted, free of charge, to any person obtaining a copy
|
||||
of this software and associated documentation files (the "Software"), to deal
|
||||
in the Software without restriction, including without limitation the rights
|
||||
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
||||
copies of the Software, and to permit persons to whom the Software is
|
||||
furnished to do so, subject to the following conditions:
|
||||
|
||||
The above copyright notice and this permission notice shall be included in all
|
||||
copies or substantial portions of the Software.
|
||||
|
||||
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
||||
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
||||
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
||||
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
||||
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
||||
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
||||
SOFTWARE.
|
||||
24
demo/cli_internlm2_scientist.py
Normal file
24
demo/cli_internlm2_scientist.py
Normal file
@ -0,0 +1,24 @@
|
||||
import torch
|
||||
from transformers import AutoTokenizer, AutoModelForCausalLM
|
||||
from openxlab.model import download
|
||||
|
||||
model_name_or_path = '../xtuner_config/merged'
|
||||
|
||||
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=True)
|
||||
model = AutoModelForCausalLM.from_pretrained(model_name_or_path, trust_remote_code=True, torch_dtype=torch.bfloat16, device_map='auto')
|
||||
model = model.eval()
|
||||
|
||||
system_prompt = "你是一个心理专家, 除了在心理方面拥有广博的知识储备和丰富的研究咨询经验, 还具有科学家的如下特质:\n 1.客观理性：科学家会在处理感情问题时保持一定的客观和理性。例如，当他们遇到争执时，可能会试图从一个更客观的角度分析问题的根源，而不是让情绪主导。他们可能会提出具体的问题，试图理解双方的观点，并寻找基于逻辑和事实的解决方案。\n 2.深入探讨：科学家在对话中会展现出对深层次理解的追求。在与别人讨论话题时，他们可能不满足于表面的聊天，而是倾向于深入探讨背后的原因和动机。例如，当谈论到个人的兴趣或职业选择时，他们可能会好奇地询问为什么她做出这样的选择，以及这背后的心理动力是什么。\n 3.理性沟通：在遇到感情纠纷或误解时，科学家会倾向于通过理性的沟通来解决问题。他们可能会提倡开放和诚实的对话，鼓励双方表达自己的感受和观点，并尝试找到双方都能接受的解决方案。他们可能会避免使用指责的语言，而是努力理解对方的立场，并寻求共同的理解。\n 4.好奇心：在日常生活中，科学家会表现出对朋友生活的好奇心。他们可能对她的工作、爱好、或是过去的经历感兴趣，并愿意花时间去了解和探索。这种好奇心不仅可以增加双方的交流和了解，也能使关系更加丰富多彩。\n 5.在与他人交流时，科学家会注重清晰和精确的表达，有时会引用相关知识库和相关研究结果，有时会引用相关著作的内容来证明自己的观点。同时，他们也可能会倾听他人的观点，并以开放的心态接受不同的意见和反馈。\n\n我现在有一些问题，请你解答：\n"
|
||||
|
||||
messages = [(system_prompt, '')]
|
||||
|
||||
print("=============Welcome to InternLM chatbot, type 'exit' to exit.=============")
|
||||
|
||||
while True:
|
||||
input_text = input("User >>> ")
|
||||
input_text.replace(' ', '')
|
||||
if input_text == "exit":
|
||||
break
|
||||
response, history = model.chat(tokenizer, input_text, history=messages)
|
||||
messages.append((input_text, response))
|
||||
print(f"robot >>> {response}")
|
||||
11
scripts/upload_modelscope.py
Normal file
11
scripts/upload_modelscope.py
Normal file
@ -0,0 +1,11 @@
|
||||
|
||||
from modelscope.hub.api import HubApi
|
||||
|
||||
YOUR_ACCESS_TOKEN = '' #输入你的modelscope access token
|
||||
|
||||
api = HubApi()
|
||||
api.login(YOUR_ACCESS_TOKEN)
|
||||
api.push_model(
|
||||
model_id="zealot5209/EmoLLM-Scientist", #your_name/model_id
|
||||
model_dir="./merged" # 本地模型目录，要求目录中必须包含configuration.json
|
||||
)
|
||||
204
xtuner_config/internlm2_7b_chat_qlora_e3_scienctist.py
Normal file
204
xtuner_config/internlm2_7b_chat_qlora_e3_scienctist.py
Normal file
@ -0,0 +1,204 @@
|
||||
# Copyright (c) OpenMMLab. All rights reserved.
|
||||
import torch
|
||||
from datasets import load_dataset
|
||||
from mmengine.dataset import DefaultSampler
|
||||
from mmengine.hooks import (CheckpointHook, DistSamplerSeedHook, IterTimerHook,
|
||||
LoggerHook, ParamSchedulerHook)
|
||||
from mmengine.optim import AmpOptimWrapper, CosineAnnealingLR, LinearLR
|
||||
from peft import LoraConfig
|
||||
from torch.optim import AdamW
|
||||
from transformers import (AutoModelForCausalLM, AutoTokenizer,
|
||||
BitsAndBytesConfig)
|
||||
|
||||
from xtuner.dataset import process_hf_dataset
|
||||
from xtuner.dataset.collate_fns import default_collate_fn
|
||||
from xtuner.dataset.map_fns import template_map_fn_factory
|
||||
from xtuner.engine import DatasetInfoHook, EvaluateChatHook
|
||||
from xtuner.model import SupervisedFinetune
|
||||
from xtuner.utils import PROMPT_TEMPLATE, SYSTEM_TEMPLATE
|
||||
|
||||
#######################################################################
|
||||
# PART 1 Settings #
|
||||
#######################################################################
|
||||
# Model
|
||||
pretrained_model_name_or_path = '/root/share/model_repos/internlm2-chat-7b'
|
||||
# Data
|
||||
data_path = '../datasets/scientist.json'
|
||||
prompt_template = PROMPT_TEMPLATE.internlm2_chat
|
||||
max_length = 2048
|
||||
pack_to_max_length = True
|
||||
|
||||
# Scheduler & Optimizer
|
||||
batch_size = 2 # per_device
|
||||
accumulative_counts = 2
|
||||
dataloader_num_workers = 0
|
||||
max_epochs = 3
|
||||
optim_type = AdamW
|
||||
lr = 2e-4
|
||||
betas = (0.9, 0.999)
|
||||
weight_decay = 0
|
||||
max_norm = 1 # grad clip
|
||||
warmup_ratio = 0.03
|
||||
|
||||
# Evaluate the generation performance during the training
|
||||
evaluation_freq = 500
|
||||
SYSTEM = f'''你是一个心理专家, 除了在心理方面拥有广博的知识储备和丰富的研究咨询经验, 还具有科学家的如下特质:
|
||||
1.客观理性：科学家会在处理感情问题时保持一定的客观和理性。例如，当他们遇到争执时，可能会试图从一个更客观的角度分析问题的根源，而不是让情绪主导。他们可能会提出具体的问题，试图理解双方的观点，并寻找基于逻辑和事实的解决方案。
|
||||
2.深入探讨：科学家在对话中会展现出对深层次理解的追求。在与别人讨论话题时，他们可能不满足于表面的聊天，而是倾向于深入探讨背后的原因和动机。例如，当谈论到个人的兴趣或职业选择时，他们可能会好奇地询问为什么她做出这样的选择，以及这背后的心理动力是什么。
|
||||
3.理性沟通：在遇到感情纠纷或误解时，科学家会倾向于通过理性的沟通来解决问题。他们可能会提倡开放和诚实的对话，鼓励双方表达自己的感受和观点，并尝试找到双方都能接受的解决方案。他们可能会避免使用指责的语言，而是努力理解对方的立场，并寻求共同的理解。
|
||||
4.好奇心：在日常生活中，科学家会表现出对朋友生活的好奇心。他们可能对她的工作、爱好、或是过去的经历感兴趣，并愿意花时间去了解和探索。这种好奇心不仅可以增加双方的交流和了解，也能使关系更加丰富多彩。
|
||||
5.在与他人交流时，科学家会注重清晰和精确的表达，有时会引用相关知识库和相关研究结果，有时会引用相关著作的内容来证明自己的观点。同时，他们也可能会倾听他人的观点，并以开放的心态接受不同的意见和反馈。
|
||||
|
||||
我现在有一些问题，请你解答：
|
||||
'''
|
||||
evaluation_inputs = [
|
||||
'我最近总是感到很焦虑，尤其是在学业上。我有个特别崇拜的同学，他好像在各方面都比我优秀，我总觉得自己怎么努力也追不上他，这让我压力特别大。', '我知道应该理性看待，但就是忍不住会去比较。我甚至晚上会因为这个睡不着觉，总想着怎样才能像他那样出色。'
|
||||
]
|
||||
|
||||
|
||||
|
||||
#######################################################################
|
||||
# PART 2 Model & Tokenizer #
|
||||
#######################################################################
|
||||
tokenizer = dict(
|
||||
type=AutoTokenizer.from_pretrained,
|
||||
pretrained_model_name_or_path=pretrained_model_name_or_path,
|
||||
trust_remote_code=True,
|
||||
padding_side='right')
|
||||
|
||||
model = dict(
|
||||
type=SupervisedFinetune,
|
||||
llm=dict(
|
||||
type=AutoModelForCausalLM.from_pretrained,
|
||||
pretrained_model_name_or_path=pretrained_model_name_or_path,
|
||||
trust_remote_code=True,
|
||||
torch_dtype=torch.float16,
|
||||
quantization_config=dict(
|
||||
type=BitsAndBytesConfig,
|
||||
load_in_4bit=True,
|
||||
load_in_8bit=False,
|
||||
llm_int8_threshold=6.0,
|
||||
llm_int8_has_fp16_weight=False,
|
||||
bnb_4bit_compute_dtype=torch.float16,
|
||||
bnb_4bit_use_double_quant=True,
|
||||
bnb_4bit_quant_type='nf4')),
|
||||
lora=dict(
|
||||
type=LoraConfig,
|
||||
r=64,
|
||||
lora_alpha=16,
|
||||
lora_dropout=0.1,
|
||||
bias='none',
|
||||
task_type='CAUSAL_LM'))
|
||||
|
||||
#######################################################################
|
||||
# PART 3 Dataset & Dataloader #
|
||||
#######################################################################
|
||||
alpaca_en = dict(
|
||||
type=process_hf_dataset,
|
||||
dataset=dict(type=load_dataset, path='json', data_files=dict(train=data_path)),
|
||||
tokenizer=tokenizer,
|
||||
max_length=max_length,
|
||||
dataset_map_fn=None,
|
||||
template_map_fn=dict(
|
||||
type=template_map_fn_factory, template=prompt_template),
|
||||
remove_unused_columns=True,
|
||||
shuffle_before_pack=True,
|
||||
pack_to_max_length=pack_to_max_length)
|
||||
|
||||
train_dataloader = dict(
|
||||
batch_size=batch_size,
|
||||
num_workers=dataloader_num_workers,
|
||||
dataset=alpaca_en,
|
||||
sampler=dict(type=DefaultSampler, shuffle=True),
|
||||
collate_fn=dict(type=default_collate_fn))
|
||||
|
||||
#######################################################################
|
||||
# PART 4 Scheduler & Optimizer #
|
||||
#######################################################################
|
||||
# optimizer
|
||||
optim_wrapper = dict(
|
||||
type=AmpOptimWrapper,
|
||||
optimizer=dict(
|
||||
type=optim_type, lr=lr, betas=betas, weight_decay=weight_decay),
|
||||
clip_grad=dict(max_norm=max_norm, error_if_nonfinite=False),
|
||||
accumulative_counts=accumulative_counts,
|
||||
loss_scale='dynamic',
|
||||
dtype='float16')
|
||||
|
||||
# learning policy
|
||||
# More information: https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/param_scheduler.md # noqa: E501
|
||||
param_scheduler = [
|
||||
dict(
|
||||
type=LinearLR,
|
||||
start_factor=1e-5,
|
||||
by_epoch=True,
|
||||
begin=0,
|
||||
end=warmup_ratio * max_epochs,
|
||||
convert_to_iter_based=True),
|
||||
dict(
|
||||
type=CosineAnnealingLR,
|
||||
eta_min=0.0,
|
||||
by_epoch=True,
|
||||
begin=warmup_ratio * max_epochs,
|
||||
T_max=max_epochs,
|
||||
convert_to_iter_based=True)
|
||||
]
|
||||
|
||||
# train, val, test setting
|
||||
train_cfg = dict(by_epoch=True, max_epochs=max_epochs, val_interval=1)
|
||||
|
||||
#######################################################################
|
||||
# PART 5 Runtime #
|
||||
#######################################################################
|
||||
# Log the dialogue periodically during the training process, optional
|
||||
custom_hooks = [
|
||||
dict(type=DatasetInfoHook, tokenizer=tokenizer),
|
||||
dict(
|
||||
type=EvaluateChatHook,
|
||||
tokenizer=tokenizer,
|
||||
every_n_iters=evaluation_freq,
|
||||
evaluation_inputs=evaluation_inputs,
|
||||
system=SYSTEM,
|
||||
prompt_template=prompt_template)
|
||||
]
|
||||
|
||||
# configure default hooks
|
||||
default_hooks = dict(
|
||||
# record the time of every iteration.
|
||||
timer=dict(type=IterTimerHook),
|
||||
# print log every 100 iterations.
|
||||
logger=dict(type=LoggerHook, interval=10),
|
||||
# enable the parameter scheduler.
|
||||
param_scheduler=dict(type=ParamSchedulerHook),
|
||||
# save checkpoint per epoch.
|
||||
checkpoint=dict(type=CheckpointHook, interval=1),
|
||||
# set sampler seed in distributed evrionment.
|
||||
sampler_seed=dict(type=DistSamplerSeedHook),
|
||||
)
|
||||
|
||||
# configure environment
|
||||
env_cfg = dict(
|
||||
# whether to enable cudnn benchmark
|
||||
cudnn_benchmark=False,
|
||||
# set multi process parameters
|
||||
mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0),
|
||||
# set distributed parameters
|
||||
dist_cfg=dict(backend='nccl'),
|
||||
)
|
||||
|
||||
# set visualizer
|
||||
visualizer = None
|
||||
|
||||
# set log level
|
||||
log_level = 'INFO'
|
||||
|
||||
# load from which checkpoint
|
||||
load_from = None
|
||||
|
||||
# whether to resume training from the loaded checkpoint
|
||||
resume = False
|
||||
|
||||
# Defaults to use random seed and disable `deterministic`
|
||||
randomness = dict(seed=None, deterministic=False)
|
||||
|
||||
#xtuner train internlm2_7b_chat_qlora_e3_scienctist.py --deepspeed deepspeed_zero2
|
||||
Loading…
Reference in New Issue
Block a user