使用 Microsoft Olive 构建您的项目
如果企业想要拥有自己的行业垂直模型，需要从数据、微调、部署开始。在之前的内容中，我们介绍了 Microsoft Olive 的内容，现在我们基于 E2E 的工作进行更详细的介绍。
Microsoft Olive 配置
如果您想了解 Microsoft Olive 的配置，请访问 使用 Microsoft Olive 进行微调。
注意 为了保持最新，请使用以下命令安装 Microsoft Olive：
pip install git+https://github.com/microsoft/Olive -U
在本地计算环境中运行 Microsoft Olive
LoRA
这个示例使用本地计算、本地数据集，并在微调文件夹中添加 olive.config
{
"input_model":{
"type": "PyTorchModel",
"config": {
"hf_config": {
"model_name": "../models/microsoft/Phi-3-mini-4k-instruct",
"task": "text-generation",
"from_pretrained_args": {
"trust_remote_code": true
}
}
}
},
"systems": {
"local_system": {
"type": "LocalSystem",
"config": {
"accelerators": [
{
"device": "GPU",
"execution_providers": [
"CUDAExecutionProvider"
]
}
]
}
}
},
"data_configs": [
{
"name": "dataset_default_train",
"type": "HuggingfaceContainer",
"params_config": {
"data_name": "json",
"data_files":"../datasets/datasets.json",
"split": "train",
"component_kwargs": {
"pre_process_data": {
"dataset_type": "corpus",
"text_cols": [
"INSTRUCTION",
"RESPONSE",
"SOURCE"
],
"text_template": "<|user|>\n{INSTRUCTION}<|end|>\n<|assistant|>\n{RESPONSE}
\n( source : {SOURCE})<|end|>",
"corpus_strategy": "join",
"source_max_len": 2048,
"pad_to_max_len": false,
"use_attention_mask": false
}
}
}
}
],
"passes": {
"lora": {
"type": "LoRA",
"config": {
"target_modules": [
"o_proj",
"qkv_proj"
],
"double_quant": true,
"lora_r": 64,
"lora_alpha": 64,
"lora_dropout": 0.1,
"train_data_config": "dataset_default_train",
"eval_dataset_size": 0.3,
"training_args": {
"seed": 0,
"data_seed": 42,
"per_device_train_batch_size": 1,
"per_device_eval_batch_size": 1,
"gradient_accumulation_steps": 4,
"gradient_checkpointing": false,
"learning_rate": 0.0001,
"num_train_epochs": 5,
"max_steps": 10,
"logging_steps": 10,
"evaluation_strategy": "steps",
"eval_steps": 187,
"group_by_length": true,
"adam_beta2": 0.999,
"max_grad_norm": 0.3
}
}
},
"merge_adapter_weights": {
"type": "MergeAdapterWeights"
},
"builder": {
"type": "ModelBuilder",
"config": {
"precision": "int4"
}
}
},
"engine": {
"log_severity_level": 0,
"host": "local_system",
"target": "local_system",
"search_strategy": false,
"cache_dir": "cache",
"output_dir" : "../models-cache/microsoft/phi3-mini-finetuned"
}
}
QLoRA
{
"input_model":{
"type": "PyTorchModel",
"config": {
"hf_config": {
"model_name": "../models/microsoft/Phi-3-mini-4k-instruct",
"task": "text-generation",
"from_pretrained_args": {
"trust_remote_code": true
}
}
}
},
"systems": {
"local_system": {
"type": "LocalSystem",
"config": {
"accelerators": [
{
"device": "GPU",
"execution_providers": [
"CUDAExecutionProvider"
]
}
]
}
}
},
"data_configs": [
{
"name": "dataset_default_train",
"type": "HuggingfaceContainer",
"params_config": {
"data_name": "json",
"data_files":"../datasets/datasets.json",
"split": "train",
"component_kwargs": {
"pre_process_data": {
"dataset_type": "corpus",
"text_cols": [
"INSTRUCTION",
"RESPONSE",
"SOURCE"
],
"text_template": "<|user|>\n{INSTRUCTION}<|end|>\n<|assistant|>\n{RESPONSE}
\n( source : {SOURCE})<|end|>",
"corpus_strategy": "join",
"source_max_len": 2048,
"pad_to_max_len": false,
"use_attention_mask": false
}
}
}
}
],
"passes": {
"qlora": {
"type": "QLoRA",
"config": {
"compute_dtype": "bfloat16",
"quant_type": "nf4",
"double_quant": true,
"lora_r": 64,
"lora_alpha": 64,
"lora_dropout": 0.1,
"train_data_config": "dataset_default_train",
"eval_dataset_size": 0.3,
"training_args": {
"seed": 0,
"data_seed": 42,
"per_device_train_batch_size": 1,
"per_device_eval_batch_size": 1,
"gradient_accumulation_steps": 4,
"gradient_checkpointing": false,
"learning_rate": 0.0001,
"num_train_epochs": 3,
"max_steps": 10,
"logging_steps": 10,
"evaluation_strategy": "steps",
"eval_steps": 187,
"group_by_length": true,
"adam_beta2": 0.999,
"max_grad_norm": 0.3
}
}
},
"merge_adapter_weights": {
"type": "MergeAdapterWeights"
},
"builder": {
"type": "ModelBuilder",
"config": {
"precision": "int4"
}
}
},
"engine": {
"log_severity_level": 0,
"host": "local_system",
"target": "local_system",
"search_strategy": false,
"cache_dir": "cache",
"output_dir" : "../models-cache/microsoft/phi3-mini-finetuned"
}
}
通知
-
如果您使用QLoRA，目前尚不支持ONNXRuntime-genai的量化转换。
-
这里需要指出的是，您可以根据自己的需求设置上述步骤。没有必要完全配置上述这些步骤。根据您的需求，您可以直接使用算法的步骤而无需微调。最后，您需要配置相关的引擎。
运行Microsoft Olive
完成Microsoft Olive后，您需要在终端运行以下命令：
olive run --config olive-config.json
通知
- 当执行Microsoft Olive时，每一步都可以被放入缓存中。我们可以通过查看微调目录来查看相关步骤的结果。
-
我们在这里提供了LoRA和QLoRA，您可以根据自己的需求进行设置。
-
推荐的运行环境是WSL/Ubuntu 22.04+。
-
为什么选择ORT？因为ORT可以部署在边缘设备上，在ORT环境中实现推理。
