update interlm2.5

README.md

@@ -46,6 +46,7 @@
 
 | 模型 | 类型 | 链接 | 模型链接 |
 | :-------------------: | :------: | :------------------------------------------------------------------------------------------------------: |:------: |
+| InternLM2_5_7B_chat | QLORA | [internlm2_5_chat_7b_qlora_oasst1_e3.py](./xtuner_config/internlm2_5_chat_7b_qlora_oasst1_e3.py) |[ModelScope](https://www.modelscope.cn/models/z342994309/emollm_interlm2_5/) |
 | InternLM2_7B_chat | QLORA | [internlm2_7b_chat_qlora_e3.py](./xtuner_config/internlm2_7b_chat_qlora_e3.py) | |
 | InternLM2_7B_chat | 全量微调 | [internlm2_chat_7b_full.py](./xtuner_config/internlm2_chat_7b_full.py) | |
 | InternLM2_7B_base | QLORA | [internlm2_7b_base_qlora_e10_M_1e4_32_64.py](./xtuner_config/internlm2_7b_base_qlora_e10_M_1e4_32_64.py) |[OpenXLab](https://openxlab.org.cn/models/detail/chg0901/EmoLLM-InternLM7B-base-10e), [ModelScope](https://www.modelscope.cn/models/chg0901/EmoLLM-InternLM7B-base-10e/summary) |
@@ -98,6 +99,7 @@
 </table>
 
 ## 🎇最近更新
+- 【2024.7】新增基于InternLM2_5_7B_chat[微调配置](./xtuner_config/internlm2_5_chat_7b_qlora_oasst1_e3.py)、模型文件发布在 [ModelScope](https://www.modelscope.cn/models/z342994309/emollm_interlm2_5/)。
 - 【2024.6】新增基于[LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory)[GLM4-9B-chat微调指南](./doc/GLM-4-9B-chat%20Lora%20微调（llama-factory）.md)、新增[基于swift的微调指南](./swift/)、论文[ESC-Eval: Evaluating Emotion Support Conversations in Large Language Models](https://arxiv.org/abs/2406.14952)引用了EmoLLM且EmoLLM取得了较好的效果。
 - 【2024.05.28】EmoLLM使用的多轮对话数据集CPsyCounD和专业评测方法已公开，详见2024 ACL findings[《CPsyCoun: A Report-based Multi-turn Dialogue Reconstruction and Evaluation Framework for Chinese Psychological Counseling》](https://arxiv.org/abs/2405.16433)!
 - 【2024.05.08】EmoLLM**爹系男友阅览体验版**上线 [1. **百度AppBuilder**](https://appbuilder.baidu.com/s/4cLyw) [2. **OpenXLab**](https://openxlab.org.cn/apps/detail/chg0901/EmoLLM3.0_Gradio_Llama3-8B-Instruct3.0), 欢迎点赞收藏
@@ -301,7 +303,8 @@ git clone https://github.com/SmartFlowAI/EmoLLM.git
 | [dream00001](https://github.com/dream00001) | 南开大学在读硕士 | | 前后端开发 |
 | [王几行XING](https://zhihu.com/people/brycewang1898) | 北京大学硕士毕业 | | 清洗数据、LLM微调、前后端开发 |
 | [思在] | 北京大学硕士毕业（微软美国） | | LLM微调、前后端开发 |
-| [TingWei](https://github.com/wwewwt) | 电子科技大学硕士毕业士 | 微信公众号：AI大模型在手 | 微调 |
+| [TingWei](https://github.com/wwewwt) | 电子科技大学硕士毕业 | 微信公众号：AI大模型在手 | 微调 |
+| [PengYu](https://github.com/hi-pengyu) | 石河子大学在读硕士 | | LLM微调 |
 ### 版权说明
 
 该项目签署了 MIT 授权许可，详情请参阅 [LICENSE](https://github.com/SmartFlowAI/EmoLLM/blob/main/LICENSE)
@@ -370,4 +373,4 @@ git clone https://github.com/SmartFlowAI/EmoLLM.git
 
 <p align="center">
  <img width="30%" src="https://private-user-images.githubusercontent.com/8240984/324394775-c8e83dac-9ed9-4a19-bb7f-b6bbedc109d9.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.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.yfBwgthq3zvmWD2givTJl5w3SMm4O5BeEFwidgG1WpY" alt="EmoLLM官方交流群">
-</p>
+</p>

README_EN.md

@@ -48,6 +48,7 @@
 
 | Model | Type | File Links | Model Links |
 | :-------------------: | :------: | :------------------------------------------------------------------------------------------------------: |:------: |
+| InternLM2_5_7B_chat | QLORA | [internlm2_5_chat_7b_qlora_oasst1_e3.py](./xtuner_config/internlm2_5_chat_7b_qlora_oasst1_e3.py) |[ModelScope](https://www.modelscope.cn/models/z342994309/emollm_interlm2_5/) |
 | InternLM2_7B_chat | QLORA | [internlm2_7b_chat_qlora_e3.py](./xtuner_config/internlm2_7b_chat_qlora_e3.py) | |
 | InternLM2_7B_chat | full fine-tuning | [internlm2_chat_7b_full.py](./xtuner_config/internlm2_chat_7b_full.py) | |
 | InternLM2_7B_base | QLORA | [internlm2_7b_base_qlora_e10_M_1e4_32_64.py](./xtuner_config/internlm2_7b_base_qlora_e10_M_1e4_32_64.py) |[OpenXLab](https://openxlab.org.cn/models/detail/chg0901/EmoLLM-InternLM7B-base-10e), [ModelScope](https://www.modelscope.cn/models/chg0901/EmoLLM-InternLM7B-base-10e/summary) |
@@ -101,6 +102,7 @@ The Model aims to fully understand and promote the mental health of individuals,
 </table>
 
 ## Recent Updates
+- 【2024.7】Added InternLM2_5_7B_chat[fine-tuning configuration](./xtuner_config/internlm2_5_chat_7b_qlora_oasst1_e3.py)、model file [ModelScope](https://www.modelscope.cn/models/z342994309/emollm_interlm2_5/)。
 - 【2024.6】 Added [LLaMA-Factory](https://github.com/hiyouga/LLaMA-Factory)[GLM4-9B-chat fine-tuning guide](./doc/GLM-4-9B-chat%20Lora%20微调（llama-factory）.md), added [swift-based fine-tuning guide](./swift/), the paper [ESC-Eval: Evaluating Emotion Support Conversations in Large Language Models](https://arxiv.org/abs/2406.14952) cited EmoLLM and EmoLLM achieved good results.
 - 【2024.05.28】The multi-turn dialogue dataset **CPsyCunD** and **professional evaluation method** used by EmoLLM have been released. For details, please see the 2024 ACL findings[《CPsyCoun: A Report-based Multi-turn Dialogue Reconstruction and Evaluation Framework for Chinese Psychological Counseling》](https://arxiv.org/abs/2405.16433)!
 - [2024.05.08] EmoLLM**Daddy-like BF V0.1** is public now in [1. **Baidu AppBuilder**](https://appbuilder.baidu.com/s/4cLyw) and [2. **OpenXLab**](https://openxlab.org.cn/apps/detail/chg0901/EmoLLM3.0_Gradio_Llama3-8B-Instruct3.0), welcome to like and add it to your collections!
@@ -304,7 +306,7 @@ This project uses Git for version control. You can see the currently available v
 | [王几行XING](zhihu.com/people/brycewang1898) | Peking University, Master's graduate | | Data Processing, LLM finetuning, Front-end and back-end development |
 | [思在] | Peking University, Master's graduate (Microsoft) | | LLM finetuning, Front-end and back-end development |
 | [TingWei](https://github.com/wwewwt) | University Of Electronic Science And Technology Of China,Master's graduate | | LLM finetuning |
-
+| [PengYu](https://github.com/hi-pengyu) | Shihezi University, Master's student | | LLM finetuning |
 ### Copyright Notice
 
 The project is licensed under the MIT License. Please refer to the details
@@ -362,4 +364,4 @@ The project is licensed under the MIT License. Please refer to the details
 
 <p align="center">
  <img width="30%" src="https://github.com/SmartFlowAI/EmoLLM/assets/62385492/55ecd0aa-4832-4269-ad57-4c26f9aa286b" alt="EmoLLM official communication group">
-</p>
+</p>

xtuner_config/internlm2_5_chat_7b_qlora_oasst1_e3.py

@@ -0,0 +1,226 @@
+# 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.hooks import (DatasetInfoHook, EvaluateChatHook,
+ VarlenAttnArgsToMessageHubHook)
+from xtuner.engine.runner import TrainLoop
+from xtuner.model import SupervisedFinetune
+from xtuner.parallel.sequence import SequenceParallelSampler
+from xtuner.utils import PROMPT_TEMPLATE
+
+#######################################################################
+# PART 1 Settings #
+#######################################################################
+# Model
+pretrained_model_name_or_path = './internlm2_5-7b-chat'
+use_varlen_attn = False
+
+# Data
+data_path = './data.jsonl'
+prompt_template = PROMPT_TEMPLATE.internlm2_chat
+max_length = 2048
+pack_to_max_length = True
+
+# parallel
+sequence_parallel_size = 1
+
+# Scheduler & Optimizer
+batch_size = 1 # per_device
+accumulative_counts = 16
+accumulative_counts *= sequence_parallel_size
+dataloader_num_workers = 0
+max_epochs = 4
+optim_type = AdamW
+lr = 2e-4
+betas = (0.9, 0.999)
+weight_decay = 0
+max_norm = 1 # grad clip
+warmup_ratio = 0.03
+
+# Save
+save_steps = 500
+save_total_limit = 2 # Maximum checkpoints to keep (-1 means unlimited)
+
+# 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,
+ use_varlen_attn=use_varlen_attn,
+ 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=32,
+ lora_alpha=64,
+ lora_dropout=0.1,
+ bias='none',
+ task_type='CAUSAL_LM'))
+
+#######################################################################
+# PART 3 Dataset & Dataloader #
+#######################################################################
+train_dataset = 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,
+ use_varlen_attn=use_varlen_attn)
+
+sampler = SequenceParallelSampler \
+ if sequence_parallel_size > 1 else DefaultSampler
+train_dataloader = dict(
+ batch_size=batch_size,
+ num_workers=dataloader_num_workers,
+ dataset=train_dataset,
+ sampler=dict(type=sampler, shuffle=True),
+ collate_fn=dict(type=default_collate_fn, use_varlen_attn=use_varlen_attn))
+
+#######################################################################
+# 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,
+ end=max_epochs,
+ convert_to_iter_based=True)
+]
+
+# train, val, test setting
+train_cfg = dict(type=TrainLoop, max_epochs=max_epochs)
+
+#######################################################################
+# 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)
+]
+
+if use_varlen_attn:
+ custom_hooks += [dict(type=VarlenAttnArgsToMessageHubHook)]
+
+# configure default hooks
+default_hooks = dict(
+ # record the time of every iteration.
+ timer=dict(type=IterTimerHook),
+ # print log every 10 iterations.
+ logger=dict(type=LoggerHook, log_metric_by_epoch=False, interval=10),
+ # enable the parameter scheduler.
+ param_scheduler=dict(type=ParamSchedulerHook),
+ # save checkpoint per `save_steps`.
+ checkpoint=dict(
+ type=CheckpointHook,
+ by_epoch=False,
+ interval=save_steps,
+ max_keep_ckpts=save_total_limit),
+ # 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)
+
+# set log processor
+log_processor = dict(by_epoch=False)