Skip to content

Commit

Permalink
Fix template encode bug (#348)
Browse files Browse the repository at this point in the history
  • Loading branch information
@Jintao-Huang
Jintao-Huang committed Jan 29, 2024
1 parent d4be1ed commit 8aaeaef
Show file tree
Hide file tree
Showing 14 changed files with 221 additions and 87 deletions.
95 changes: 67 additions & 28 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -148,41 +148,80 @@ Here is a simple introduction of web-ui:
- View the training time and training GPU memory comparison under different parameters, you can check [Benchmark](https://github.com/modelscope/swift/blob/main/docs/source/LLM/Benchmark.md).


### Quick Start
```python
# pip install ms-swift -U

# Experimental environment: A10, 3090, V100, ...
# 12GB GPU memory
import os
os.environ['CUDA_VISIBLE_DEVICES'] = '0'

import torch

from swift.llm import (
DatasetName, InferArguments, ModelType, SftArguments,
infer_main, sft_main, app_ui_main, merge_lora_main
)

model_type = ModelType.qwen_1_8b_chat
sft_args = SftArguments(
model_type=model_type,
train_dataset_sample=2000,
dataset=[DatasetName.jd_sentiment_zh],
output_dir='output')
result = sft_main(sft_args)
best_model_checkpoint = result['best_model_checkpoint']
print(f'best_model_checkpoint: {best_model_checkpoint}')
torch.cuda.empty_cache()

infer_args = InferArguments(
ckpt_dir=best_model_checkpoint,
load_dataset_config=True,
show_dataset_sample=10)
# merge_lora_main(infer_args)
result = infer_main(infer_args)
torch.cuda.empty_cache()

app_ui_main(infer_args)
```


### Features
- Supported SFT Methods: [lora](https://arxiv.org/abs/2106.09685), [qlora](https://arxiv.org/abs/2305.14314), [longlora](https://arxiv.org/abs/2309.12307), [qalora](https://arxiv.org/abs/2309.14717), full parameter fine-tuning, partial parameter fine-tuning.
- Supported Features: quantization, DDP, model parallelism, gradient checkpointing, pushing to modelscope hub, custom datasets, multimodal and agent SFT, mutli-round chat, DPO, self-cognition fine-tuning, ...
- Supported Models: [[Detailed Info]](https://github.com/modelscope/swift/blob/main/docs/source/LLM/%E6%94%AF%E6%8C%81%E7%9A%84%E6%A8%A1%E5%9E%8B%E5%92%8C%E6%95%B0%E6%8D%AE%E9%9B%86.md#%E6%A8%A1%E5%9E%8B)
- Multi-Modal:
- qwen-vl series: qwen-vl, qwen-vl-chat, qwen-vl-chat-int4.
- qwen-audio series: qwen-audio, qwen-audio-chat.
- yi-vl series: yi-vl-6b-chat, yi-vl-34b-chat.
- cogagent series: cogagent-18b-chat, cogagent-18b-instruct.
- [qwen-vl](https://github.com/QwenLM/Qwen-VL) series: qwen-vl, qwen-vl-chat, qwen-vl-chat-int4.
- [qwen-audio](https://github.com/QwenLM/Qwen-Audio) series: qwen-audio, qwen-audio-chat.
- [yi-vl](https://github.com/01-ai/Yi) series: yi-vl-6b-chat, yi-vl-34b-chat.
- [cogagent](https://github.com/THUDM/CogVLM) series: cogagent-18b-chat, cogagent-18b-instruct.
- General:
- qwen series: qwen-1_8b, qwen-1_8b-chat, qwen-1_8b-chat-int4, qwen-1_8b-chat-int8, qwen-7b, qwen-7b-chat, qwen-7b-chat-int4, qwen-7b-chat-int8, qwen-14b, qwen-14b-chat, qwen-14b-chat-int4, qwen-14b-chat-int8, qwen-72b, qwen-72b-chat, qwen-72b-chat-int4, qwen-72b-chat-int8.
- chatglm series: chatglm2-6b, chatglm2-6b-32k, chatglm3-6b-base, chatglm3-6b, chatglm3-6b-32k.
- llama series: llama2-7b, llama2-7b-chat, llama2-13b, llama2-13b-chat, llama2-70b, llama2-70b-chat.
- yi series: yi-6b, yi-6b-200k, yi-6b-chat, yi-34b, yi-34b-200k, yi-34b-chat.
- internlm series: internlm-7b, internlm-7b-chat, internlm-7b-chat-8k, internlm-20b, internlm-20b-chat, internlm2-7b-base, internlm2-7b, internlm2-7b-sft-chat, internlm2-7b-chat, internlm2-20b-base, internlm2-20b, internlm2-20b-sft-chat, internlm2-20b-chat.
- deepseek series: deepseek-7b, deepseek-7b-chat, deepseek-67b, deepseek-67b-chat, deepseek-moe-16b, deepseek-moe-16b-chat.
- openbuddy series: openbuddy-llama2-13b-chat, openbuddy-llama-65b-chat, openbuddy-llama2-70b-chat, openbuddy-mistral-7b-chat, openbuddy-zephyr-7b-chat, openbuddy-deepseek-67b-chat.
- mistral series: mistral-7b, mistral-7b-instruct, mistral-7b-instruct-v2.
- mixtral series: mixtral-moe-7b, mixtral-moe-7b-instruct.
- baichuan series: baichuan-7b, baichuan-13b, baichuan-13b-chat, baichuan2-7b, baichuan2-7b-chat, baichuan2-13b, baichuan2-13b-chat, baichuan2-7b-chat-int4, baichuan2-13b-chat-int4.
- yuan series: yuan2-2b-instruct, yuan2-2b-janus-instruct, yuan2-51b-instruct, yuan2-102b-instruct.
- xverse series: xverse-7b, xverse-7b-chat, xverse-13b, xverse-13b-chat, xverse-65b, xverse-65b-v2, xverse-65b-chat, xverse-13b-256k.
- orion series: orion-14b, orion-14b-chat.
- bluelm series: bluelm-7b, bluelm-7b-chat, bluelm-7b-32k, bluelm-7b-chat-32k.
- zephyr series: zephyr-7b-beta-chat.
- ziya series: ziya2-13b, ziya2-13b-chat.
- skywork series: skywork-13b, skywork-13b-chat.
- other: polylm-13b, seqgpt-560m, sus-34b-chat.
- [qwen](https://github.com/QwenLM/Qwen) series: qwen-1_8b, qwen-1_8b-chat, qwen-1_8b-chat-int4, qwen-1_8b-chat-int8, qwen-7b, qwen-7b-chat, qwen-7b-chat-int4, qwen-7b-chat-int8, qwen-14b, qwen-14b-chat, qwen-14b-chat-int4, qwen-14b-chat-int8, qwen-72b, qwen-72b-chat, qwen-72b-chat-int4, qwen-72b-chat-int8.
- [chatglm](https://github.com/THUDM/ChatGLM-6B) series: chatglm2-6b, chatglm2-6b-32k, chatglm3-6b-base, chatglm3-6b, chatglm3-6b-32k.
- [llama](https://github.com/facebookresearch/llama) series: llama2-7b, llama2-7b-chat, llama2-13b, llama2-13b-chat, llama2-70b, llama2-70b-chat.
- [yi](https://github.com/01-ai/Yi) series: yi-6b, yi-6b-200k, yi-6b-chat, yi-34b, yi-34b-200k, yi-34b-chat.
- [internlm](https://github.com/InternLM/InternLM) series: internlm-7b, internlm-7b-chat, internlm-7b-chat-8k, internlm-20b, internlm-20b-chat, internlm2-7b-base, internlm2-7b, internlm2-7b-sft-chat, internlm2-7b-chat, internlm2-20b-base, internlm2-20b, internlm2-20b-sft-chat, internlm2-20b-chat.
- [deepseek](https://github.com/deepseek-ai/deepseek-LLM) series: deepseek-7b, deepseek-7b-chat, deepseek-67b, deepseek-67b-chat, deepseek-moe-16b, deepseek-moe-16b-chat.
- [openbuddy](https://github.com/OpenBuddy/OpenBuddy) series: openbuddy-llama2-13b-chat, openbuddy-llama-65b-chat, openbuddy-llama2-70b-chat, openbuddy-mistral-7b-chat, openbuddy-zephyr-7b-chat, openbuddy-deepseek-67b-chat.
- [mistral](https://github.com/mistralai/mistral-src) series: mistral-7b, mistral-7b-instruct, mistral-7b-instruct-v2.
- [mixtral](https://github.com/mistralai/mistral-src) series: mixtral-moe-7b, mixtral-moe-7b-instruct.
- [baichuan](https://github.com/baichuan-inc/Baichuan2) series: baichuan-7b, baichuan-13b, baichuan-13b-chat, baichuan2-7b, baichuan2-7b-chat, baichuan2-13b, baichuan2-13b-chat, baichuan2-7b-chat-int4, baichuan2-13b-chat-int4.
- [yuan](https://github.com/IEIT-Yuan/Yuan-2.0) series: yuan2-2b-instruct, yuan2-2b-janus-instruct, yuan2-51b-instruct, yuan2-102b-instruct.
- [xverse](https://github.com/xverse-ai/XVERSE-13B) series: xverse-7b, xverse-7b-chat, xverse-13b, xverse-13b-chat, xverse-65b, xverse-65b-v2, xverse-65b-chat, xverse-13b-256k.
- [orion](https://github.com/OrionStarAI/OrionStar-Yi-34B-Chat) series: orion-14b, orion-14b-chat.
- [bluelm](https://github.com/vivo-ai-lab/BlueLM) series: bluelm-7b, bluelm-7b-chat, bluelm-7b-32k, bluelm-7b-chat-32k.
- [zephyr](https://github.com/huggingface/alignment-handbook) series: zephyr-7b-beta-chat.
- [ziya](https://github.com/IDEA-CCNL/Fengshenbang-LM) series: ziya2-13b, ziya2-13b-chat.
- [skywork](https://github.com/SkyworkAI/Skywork) series: skywork-13b, skywork-13b-chat.
- other: [polylm-13b](https://github.com/DAMO-NLP-MT/PolyLM), [seqgpt-560m](https://github.com/Alibaba-NLP/SeqGPT), [sus-34b-chat](https://github.com/SUSTech-IDEA/SUS-Chat).
- Financial:
- tongyi-finance series: tongyi-finance-14b, tongyi-finance-14b-chat, tongyi-finance-14b-chat-int4.
- [tongyi-finance](https://github.com/QwenLM/Qwen) series: tongyi-finance-14b, tongyi-finance-14b-chat, tongyi-finance-14b-chat-int4.
- Coding:
- codefuse series: codefuse-codellama-34b-chat, codefuse-codegeex2-6b-chat, codefuse-qwen-14b-chat.
- deepseek-coder series: deepseek-coder-1_3b, deepseek-coder-1_3b-instruct, deepseek-coder-6_7b, deepseek-coder-6_7b-instruct, deepseek-coder-33b, deepseek-coder-33b-instruct.
- codegeex2 series: codegeex2-6b.
- phi series: phi2-3b.
- [codefuse](https://github.com/codefuse-ai) series: codefuse-codellama-34b-chat, codefuse-codegeex2-6b-chat, codefuse-qwen-14b-chat.
- [deepseek-coder](https://github.com/deepseek-ai/DeepSeek-Coder) series: deepseek-coder-1_3b, deepseek-coder-1_3b-instruct, deepseek-coder-6_7b, deepseek-coder-6_7b-instruct, deepseek-coder-33b, deepseek-coder-33b-instruct.
- [codegeex2](https://github.com/THUDM/CodeGeeX2) series: codegeex2-6b.
- [phi](https://www.microsoft.com/en-us/research/blog/phi-2-the-surprising-power-of-small-language-models/) series: phi2-3b.
- Supported Datasets: [[Detailed Info]](https://github.com/modelscope/swift/blob/main/docs/source/LLM/%E6%94%AF%E6%8C%81%E7%9A%84%E6%A8%A1%E5%9E%8B%E5%92%8C%E6%95%B0%E6%8D%AE%E9%9B%86.md#%E6%95%B0%E6%8D%AE%E9%9B%86)
- NLP:
- General: 🔥alpaca-en(gpt4), 🔥alpaca-zh(gpt4), multi-alpaca-all, instinwild-en, instinwild-zh, cot-en, cot-zh, firefly-all-zh, instruct-en, gpt4all-en, sharegpt-en, sharegpt-zh, tutu-v2-sft-mixture, wikipedia-zh, open-orca, open-orca-gpt4, sharegpt-gpt4.
Expand All @@ -202,7 +241,7 @@ Here is a simple introduction of web-ui:
- Custom Dataset
- Supported Templates:
- Text Generation: default-generation, default-generation-bos, chatglm-generation.
- Chat: default, qwen, baichuan, chatglm2, chatglm3, llama, openbuddy, internlm, internlm2, yi, yuan, xverse, ziya, skywork, bluelm, zephyr, sus, deepseek, deepseek-coder, codefuse-codellama, codefuse, cogagent-chat, cogagent-instruct.
- Chat: default, qwen, baichuan, chatglm2, chatglm3, llama, openbuddy, internlm, internlm2, yi, yuan, xverse, ziya, skywork, bluelm, zephyr, sus, deepseek, deepseek-coder, codefuse-codellama, codefuse, cogagent-chat, cogagent-instruct, yi-vl.


## 🔥SCEdit
Expand Down
95 changes: 67 additions & 28 deletions README_CN.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -148,41 +148,80 @@ swift web-ui
- 查看不同参数下的训练时间和训练显存对比, 可以查看[Benchmark](https://github.com/modelscope/swift/blob/main/docs/source/LLM/Benchmark.md).


### 快速开始
```python
# pip install ms-swift -U

# Experimental environment: A10, 3090, V100, ...
# 12GB GPU memory
import os
os.environ['CUDA_VISIBLE_DEVICES'] = '0'

import torch

from swift.llm import (
DatasetName, InferArguments, ModelType, SftArguments,
infer_main, sft_main, app_ui_main, merge_lora_main
)

model_type = ModelType.qwen_1_8b_chat
sft_args = SftArguments(
model_type=model_type,
train_dataset_sample=2000,
dataset=[DatasetName.jd_sentiment_zh],
output_dir='output')
result = sft_main(sft_args)
best_model_checkpoint = result['best_model_checkpoint']
print(f'best_model_checkpoint: {best_model_checkpoint}')
torch.cuda.empty_cache()

infer_args = InferArguments(
ckpt_dir=best_model_checkpoint,
load_dataset_config=True,
show_dataset_sample=10)
# merge_lora_main(infer_args)
result = infer_main(infer_args)
torch.cuda.empty_cache()

app_ui_main(infer_args)
```


### 特性
- 支持的SFT方法: [lora](https://arxiv.org/abs/2106.09685), [qlora](https://arxiv.org/abs/2305.14314), [longlora](https://arxiv.org/abs/2309.12307), [qalora](https://arxiv.org/abs/2309.14717), 全参数微调, 部分参数微调.
- 支持的特性: 模型量化, DDP, 模型并行, gradient checkpointing, 支持推送ModelScope Hub, 自定义数据集, 多模态和Agent SFT, 多轮对话, DPO, 自我认知微调, ...
- 支持的模型: [[详细信息]](https://github.com/modelscope/swift/blob/main/docs/source/LLM/%E6%94%AF%E6%8C%81%E7%9A%84%E6%A8%A1%E5%9E%8B%E5%92%8C%E6%95%B0%E6%8D%AE%E9%9B%86.md#%E6%A8%A1%E5%9E%8B)
- 多模态:
- qwen-vl 系列: qwen-vl, qwen-vl-chat, qwen-vl-chat-int4.
- qwen-audio 系列: qwen-audio, qwen-audio-chat.
- yi-vl 系列: yi-vl-6b-chat, yi-vl-34b-chat.
- cogagent 系列: cogagent-18b-chat, cogagent-18b-instruct.
- [qwen-vl](https://github.com/QwenLM/Qwen-VL) 系列: qwen-vl, qwen-vl-chat, qwen-vl-chat-int4.
- [qwen-audio](https://github.com/QwenLM/Qwen-Audio) 系列: qwen-audio, qwen-audio-chat.
- [yi-vl](https://github.com/01-ai/Yi) 系列: yi-vl-6b-chat, yi-vl-34b-chat.
- [cogagent](https://github.com/THUDM/CogVLM) 系列: cogagent-18b-chat, cogagent-18b-instruct.
- 通用:
- qwen 系列: qwen-1_8b, qwen-1_8b-chat, qwen-1_8b-chat-int4, qwen-1_8b-chat-int8, qwen-7b, qwen-7b-chat, qwen-7b-chat-int4, qwen-7b-chat-int8, qwen-14b, qwen-14b-chat, qwen-14b-chat-int4, qwen-14b-chat-int8, qwen-72b, qwen-72b-chat, qwen-72b-chat-int4, qwen-72b-chat-int8.
- chatglm 系列: chatglm2-6b, chatglm2-6b-32k, chatglm3-6b-base, chatglm3-6b, chatglm3-6b-32k.
- llama 系列: llama2-7b, llama2-7b-chat, llama2-13b, llama2-13b-chat, llama2-70b, llama2-70b-chat.
- yi 系列: yi-6b, yi-6b-200k, yi-6b-chat, yi-34b, yi-34b-200k, yi-34b-chat.
- internlm 系列: internlm-7b, internlm-7b-chat, internlm-7b-chat-8k, internlm-20b, internlm-20b-chat, internlm2-7b-base, internlm2-7b, internlm2-7b-sft-chat, internlm2-7b-chat, internlm2-20b-base, internlm2-20b, internlm2-20b-sft-chat, internlm2-20b-chat.
- deepseek 系列: deepseek-7b, deepseek-7b-chat, deepseek-67b, deepseek-67b-chat, deepseek-moe-16b, deepseek-moe-16b-chat.
- openbuddy 系列: openbuddy-llama2-13b-chat, openbuddy-llama-65b-chat, openbuddy-llama2-70b-chat, openbuddy-mistral-7b-chat, openbuddy-zephyr-7b-chat, openbuddy-deepseek-67b-chat.
- mistral 系列: mistral-7b, mistral-7b-instruct, mistral-7b-instruct-v2.
- mixtral 系列: mixtral-moe-7b, mixtral-moe-7b-instruct.
- baichuan 系列: baichuan-7b, baichuan-13b, baichuan-13b-chat, baichuan2-7b, baichuan2-7b-chat, baichuan2-13b, baichuan2-13b-chat, baichuan2-7b-chat-int4, baichuan2-13b-chat-int4.
- yuan 系列: yuan2-2b-instruct, yuan2-2b-janus-instruct, yuan2-51b-instruct, yuan2-102b-instruct.
- xverse 系列: xverse-7b, xverse-7b-chat, xverse-13b, xverse-13b-chat, xverse-65b, xverse-65b-v2, xverse-65b-chat, xverse-13b-256k.
- orion 系列: orion-14b, orion-14b-chat.
- bluelm 系列: bluelm-7b, bluelm-7b-chat, bluelm-7b-32k, bluelm-7b-chat-32k.
- zephyr 系列: zephyr-7b-beta-chat.
- ziya 系列: ziya2-13b, ziya2-13b-chat.
- skywork 系列: skywork-13b, skywork-13b-chat.
- other: polylm-13b, seqgpt-560m, sus-34b-chat.
- [qwen](https://github.com/QwenLM/Qwen) 系列: qwen-1_8b, qwen-1_8b-chat, qwen-1_8b-chat-int4, qwen-1_8b-chat-int8, qwen-7b, qwen-7b-chat, qwen-7b-chat-int4, qwen-7b-chat-int8, qwen-14b, qwen-14b-chat, qwen-14b-chat-int4, qwen-14b-chat-int8, qwen-72b, qwen-72b-chat, qwen-72b-chat-int4, qwen-72b-chat-int8.
- [chatglm](https://github.com/THUDM/ChatGLM-6B) 系列: chatglm2-6b, chatglm2-6b-32k, chatglm3-6b-base, chatglm3-6b, chatglm3-6b-32k.
- [llama](https://github.com/facebookresearch/llama) 系列: llama2-7b, llama2-7b-chat, llama2-13b, llama2-13b-chat, llama2-70b, llama2-70b-chat.
- [yi](https://github.com/01-ai/Yi) 系列: yi-6b, yi-6b-200k, yi-6b-chat, yi-34b, yi-34b-200k, yi-34b-chat.
- [internlm](https://github.com/InternLM/InternLM) 系列: internlm-7b, internlm-7b-chat, internlm-7b-chat-8k, internlm-20b, internlm-20b-chat, internlm2-7b-base, internlm2-7b, internlm2-7b-sft-chat, internlm2-7b-chat, internlm2-20b-base, internlm2-20b, internlm2-20b-sft-chat, internlm2-20b-chat.
- [deepseek](https://github.com/deepseek-ai/deepseek-LLM) 系列: deepseek-7b, deepseek-7b-chat, deepseek-67b, deepseek-67b-chat, deepseek-moe-16b, deepseek-moe-16b-chat.
- [openbuddy](https://github.com/OpenBuddy/OpenBuddy) 系列: openbuddy-llama2-13b-chat, openbuddy-llama-65b-chat, openbuddy-llama2-70b-chat, openbuddy-mistral-7b-chat, openbuddy-zephyr-7b-chat, openbuddy-deepseek-67b-chat.
- [mistral](https://github.com/mistralai/mistral-src) 系列: mistral-7b, mistral-7b-instruct, mistral-7b-instruct-v2.
- [mixtral](https://github.com/mistralai/mistral-src) 系列: mixtral-moe-7b, mixtral-moe-7b-instruct.
- [baichuan](https://github.com/baichuan-inc/Baichuan2) 系列: baichuan-7b, baichuan-13b, baichuan-13b-chat, baichuan2-7b, baichuan2-7b-chat, baichuan2-13b, baichuan2-13b-chat, baichuan2-7b-chat-int4, baichuan2-13b-chat-int4.
- [yuan](https://github.com/IEIT-Yuan/Yuan-2.0) 系列: yuan2-2b-instruct, yuan2-2b-janus-instruct, yuan2-51b-instruct, yuan2-102b-instruct.
- [xverse](https://github.com/xverse-ai/XVERSE-13B) 系列: xverse-7b, xverse-7b-chat, xverse-13b, xverse-13b-chat, xverse-65b, xverse-65b-v2, xverse-65b-chat, xverse-13b-256k.
- [orion](https://github.com/OrionStarAI/OrionStar-Yi-34B-Chat) 系列: orion-14b, orion-14b-chat.
- [bluelm](https://github.com/vivo-ai-lab/BlueLM) 系列: bluelm-7b, bluelm-7b-chat, bluelm-7b-32k, bluelm-7b-chat-32k.
- [zephyr](https://github.com/huggingface/alignment-handbook) 系列: zephyr-7b-beta-chat.
- [ziya](https://github.com/IDEA-CCNL/Fengshenbang-LM) 系列: ziya2-13b, ziya2-13b-chat.
- [skywork](https://github.com/SkyworkAI/Skywork) 系列: skywork-13b, skywork-13b-chat.
- other: [polylm-13b](https://github.com/DAMO-NLP-MT/PolyLM), [seqgpt-560m](https://github.com/Alibaba-NLP/SeqGPT), [sus-34b-chat](https://github.com/SUSTech-IDEA/SUS-Chat).
- 金融:
- tongyi-finance 系列: tongyi-finance-14b, tongyi-finance-14b-chat, tongyi-finance-14b-chat-int4.
- [tongyi-finance](https://github.com/QwenLM/Qwen) 系列: tongyi-finance-14b, tongyi-finance-14b-chat, tongyi-finance-14b-chat-int4.
- 代码:
- codefuse 系列: codefuse-codellama-34b-chat, codefuse-codegeex2-6b-chat, codefuse-qwen-14b-chat.
- deepseek-coder 系列: deepseek-coder-1_3b, deepseek-coder-1_3b-instruct, deepseek-coder-6_7b, deepseek-coder-6_7b-instruct, deepseek-coder-33b, deepseek-coder-33b-instruct.
- codegeex2 系列: codegeex2-6b.
- phi 系列: phi2-3b.
- [codefuse](https://github.com/codefuse-ai) 系列: codefuse-codellama-34b-chat, codefuse-codegeex2-6b-chat, codefuse-qwen-14b-chat.
- [deepseek-coder](https://github.com/deepseek-ai/DeepSeek-Coder) 系列: deepseek-coder-1_3b, deepseek-coder-1_3b-instruct, deepseek-coder-6_7b, deepseek-coder-6_7b-instruct, deepseek-coder-33b, deepseek-coder-33b-instruct.
- [codegeex2](https://github.com/THUDM/CodeGeeX2) 系列: codegeex2-6b.
- [phi](https://www.microsoft.com/en-us/research/blog/phi-2-the-surprising-power-of-small-language-models/) 系列: phi2-3b.
- 支持的数据集: [[详细信息]](https://github.com/modelscope/swift/blob/main/docs/source/LLM/%E6%94%AF%E6%8C%81%E7%9A%84%E6%A8%A1%E5%9E%8B%E5%92%8C%E6%95%B0%E6%8D%AE%E9%9B%86.md#%E6%95%B0%E6%8D%AE%E9%9B%86)
- NLP:
- 通用: 🔥alpaca-en(gpt4), 🔥alpaca-zh(gpt4), multi-alpaca-all, instinwild-en, instinwild-zh, cot-en, cot-zh, firefly-all-zh, instruct-en, gpt4all-en, sharegpt-en, sharegpt-zh, tutu-v2-sft-mixture, wikipedia-zh, open-orca, open-orca-gpt4, sharegpt-gpt4.
Expand All @@ -202,7 +241,7 @@ swift web-ui
- 自定义数据集
- 支持的对话模板:
- 文本生成: default-generation, default-generation-bos, chatglm-generation.
- 对话: default, qwen, baichuan, chatglm2, chatglm3, llama, openbuddy, internlm, internlm2, yi, yuan, xverse, ziya, skywork, bluelm, zephyr, sus, deepseek, deepseek-coder, codefuse-codellama, codefuse, cogagent-chat, cogagent-instruct.
- 对话: default, qwen, baichuan, chatglm2, chatglm3, llama, openbuddy, internlm, internlm2, yi, yuan, xverse, ziya, skywork, bluelm, zephyr, sus, deepseek, deepseek-coder, codefuse-codellama, codefuse, cogagent-chat, cogagent-instruct, yi-vl.


## 🔥SCEdit
Expand Down
1 change: 1 addition & 0 deletions docs/source/LLM/LLM微调文档.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -63,6 +63,7 @@ torch.cuda.empty_cache()

infer_args = InferArguments(
ckpt_dir=best_model_checkpoint,
load_dataset_config=True,
show_dataset_sample=10)
# merge_lora_main(infer_args)
result = infer_main(infer_args)
Expand Down
Loading

0 comments on commit 8aaeaef

Please sign in to comment.