File size: 3,009 Bytes
f677bb1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
# 导出与推送


## Merge LoRA

- 查看[这里](https://github.com/modelscope/ms-swift/blob/main/examples/export/merge_lora.sh)。

## 量化

SWFIT支持AWQ、GPTQ、BNB模型的量化导出。其中使用AWQ、GPTQ需使用校准数据集,量化性能较好但量化耗时较长;而BNB无需校准数据集,量化耗时较短。

| 量化技术 | 多模态 | 推理加速 | 继续训练 |
| -------- | ------ | -------- | -------- |
| GPTQ     | ✅      | ✅        | ✅        |
| AWQ      | ✅      | ✅        | ✅        |
| BNB      | ❌      | ✅        | ✅        |


除SWIFT安装外,需要安装以下额外依赖:
```shell
# 使用awq量化:
# autoawq和cuda版本有对应关系,请按照`https://github.com/casper-hansen/AutoAWQ`选择版本
# 如果出现torch依赖冲突,请额外增加指令`--no-deps`
pip install autoawq -U

# 使用gptq量化:
# auto_gptq和cuda版本有对应关系,请按照`https://github.com/PanQiWei/AutoGPTQ#quick-installation`选择版本
pip install auto_gptq optimum -U

# 使用bnb量化:
pip install bitsandbytes -U
```

我们提供了一系列脚本展现SWIFT的量化导出能力:
- 支持[AWQ](https://github.com/modelscope/ms-swift/blob/main/examples/export/quantize/awq.sh)/[GPTQ](https://github.com/modelscope/ms-swift/blob/main/examples/export/quantize/gptq.sh)/[BNB](https://github.com/modelscope/ms-swift/blob/main/examples/export/quantize/bnb.sh)量化导出。
- 多模态量化: 支持使用GPTQ和AWQ对多模态模型进行量化,其中AWQ支持的多模态模型有限。参考[这里](https://github.com/modelscope/ms-swift/tree/main/examples/export/quantize/mllm)。
- 更多系列模型的支持: 支持[Bert](https://github.com/modelscope/ms-swift/tree/main/examples/export/quantize/bert),[Reward Model](https://github.com/modelscope/ms-swift/tree/main/examples/export/quantize/reward_model)的量化导出。
- 使用SWIFT量化导出的模型支持使用vllm/lmdeploy进行推理加速;也支持使用QLoRA继续进行SFT/RLHF。


## 推送模型

SWIFT支持将训练/量化的模型重新推送到ModelScope/HuggingFace。默认推送到ModelScope,你可以指定`--use_hf true`推送到HuggingFace。
```shell
swift export \
    --model output/vx-xxx/checkpoint-xxx \
    --push_to_hub true \
    --hub_model_id '<model-id>' \
    --hub_token '<sdk-token>' \
    --use_hf false
```

小贴士:
- 你可以使用`--model <checkpoint-dir>`或者`--adapters <checkpoint-dir>`指定需要推送的checkpoint目录,这两种写法在推送模型场景没有差异。
- 推送到ModelScope时,你需要确保你已经注册了魔搭账号,你的SDK token可以在[该页面](https://www.modelscope.cn/my/myaccesstoken)中获取。推送模型需确保sdk token的账号具有model_id对应组织的编辑权限。推送模型将自动创建对应model_id的模型仓库(如果该模型仓库不存在),你可以使用`--hub_private_repo true`来自动创建私有的模型仓库。