LLM: add qlora finetuning example using trl.SFTTrainer (#10183)

python/llm/example/GPU/LLM-Finetuning/QLoRA/README.md

@@ -2,4 +2,4 @@
 
 We provide [Alpaca-QLoRA example](./alpaca-qlora/), which ports [Alpaca-LoRA](https://github.com/tloen/alpaca-lora/tree/main) to BigDL-LLM (using [QLoRA](https://arxiv.org/abs/2305.14314) algorithm) on [Intel GPU](../../README.md).
 
-Meanwhile, we also provide a [simple example](./simple-example/) to help you get started with QLoRA Finetuning using BigDL-LLM.
+Meanwhile, we also provide a [simple example](./simple-example/) to help you get started with QLoRA Finetuning using BigDL-LLM, and [TRL example](./trl-example/) to help you get started with QLoRA Finetuning using BigDL-LLM and TRL library.

python/llm/example/GPU/LLM-Finetuning/QLoRA/trl-example/README.md

@@ -0,0 +1,55 @@
+# Example of QLoRA Finetuning with BigDL-LLM
+
+This simple example demonstrates how to finetune a llama2-7b model use BigDL-LLM 4bit optimizations with TRL library on [Intel GPU](../../../README.md).
+Note, this example is just used for illustrating related usage and don't guarantee convergence of training.
+
+## 0. Requirements
+To run this example with BigDL-LLM on Intel GPUs, we have some recommended requirements for your machine, please refer to [here](../../../README.md#requirements) for more information.
+
+## Example: Finetune llama2-7b using qlora
+
+The `export_merged_model.py` is ported from [alpaca-lora](https://github.com/tloen/alpaca-lora/blob/main/export_hf_checkpoint.py).
+
+### 1. Install
+
+```bash
+conda create -n llm python=3.9
+conda activate llm
+# below command will install intel_extension_for_pytorch==2.1.10+xpu as default
+pip install --pre --upgrade bigdl-llm[xpu] -f https://developer.intel.com/ipex-whl-stable-xpu
+pip install transformers==4.34.0 datasets
+pip install peft==0.5.0
+pip install accelerate==0.23.0
+pip install bitsandbytes scipy trl
+```
+
+### 2. Configures OneAPI environment variables
+```bash
+source /opt/intel/oneapi/setvars.sh
+```
+
+### 3. Finetune model
+
+```
+python ./qlora_finetuning.py --repo-id-or-model-path REPO_ID_OR_MODEL_PATH
+```
+
+#### Sample Output
+```log
+{'loss': 1.7386, 'learning_rate': 8.888888888888888e-06, 'epoch': 0.19}            
+{'loss': 1.9242, 'learning_rate': 6.666666666666667e-06, 'epoch': 0.22}            
+{'loss': 1.6819, 'learning_rate': 4.444444444444444e-06, 'epoch': 0.26}            
+{'loss': 1.755, 'learning_rate': 2.222222222222222e-06, 'epoch': 0.29}             
+{'loss': 1.7455, 'learning_rate': 0.0, 'epoch': 0.32}                              
+{'train_runtime': 172.8523, 'train_samples_per_second': 4.628, 'train_steps_per_second': 1.157, 'train_loss': 1.9101631927490235, 'epoch': 0.32}
+100%|████████████████████████████████████████████| 200/200 [02:52<00:00,  1.16it/s]
+TrainOutput(global_step=200, training_loss=1.9101631927490235, metrics={'train_runtime': 172.8523, 'train_samples_per_second': 4.628, 'train_steps_per_second': 1.157, 'train_loss': 1.9101631927490235, 'epoch': 0.32})
+```
+
+### 4. Merge the adapter into the original model
+
+```
+python ./export_merged_model.py --repo-id-or-model-path REPO_ID_OR_MODEL_PATH --adapter_path ./outputs/checkpoint-200 --output_path ./outputs/checkpoint-200-merged
+```
+
+Then you can use `./outputs/checkpoint-200-merged` as a normal huggingface transformer model to do inference.

python/llm/example/GPU/LLM-Finetuning/QLoRA/trl-example/export_merged_model.py

@@ -0,0 +1,44 @@
+#
+# Copyright 2016 The BigDL Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+import os
+
+import torch
+from transformers import LlamaTokenizer  # noqa: F402
+import argparse
+
+current_dir = os.path.dirname(os.path.realpath(__file__))
+common_util_path = os.path.join(current_dir, '..', '..')
+import sys
+sys.path.append(common_util_path)
+from common.utils import merge_adapter
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser(description='Merge the adapter into the original model for Llama2 model')
+    parser.add_argument('--repo-id-or-model-path', type=str, default="meta-llama/Llama-2-7b-hf",
+                        help='The huggingface repo id for the Llama2 (e.g. `meta-llama/Llama-2-7b-hf` and `meta-llama/Llama-2-13b-chat-hf`) to be downloaded'
+                             ', or the path to the huggingface checkpoint folder')
+    parser.add_argument('--adapter_path', type=str,)
+    parser.add_argument('--output_path', type=str,)
+
+    args = parser.parse_args()
+    base_model = model_path = args.repo_id_or_model_path
+    adapter_path = args.adapter_path
+    output_path = args.output_path
+    
+    tokenizer = LlamaTokenizer.from_pretrained(base_model)
+    merge_adapter(base_model, tokenizer, adapter_path, output_path)
+    print(f'Finish to merge the adapter into the original model and you could find the merged model in {output_path}.')

python/llm/example/GPU/LLM-Finetuning/QLoRA/trl-example/qlora_finetuning.py

@@ -0,0 +1,94 @@
+#
+# Copyright 2016 The BigDL Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+import torch
+import os
+
+import transformers
+from transformers import LlamaTokenizer
+from peft import LoraConfig
+from transformers import BitsAndBytesConfig
+from bigdl.llm.transformers.qlora import get_peft_model, prepare_model_for_kbit_training
+from bigdl.llm.transformers import AutoModelForCausalLM
+from datasets import load_dataset
+from trl import SFTTrainer
+import argparse
+
+if __name__ == "__main__":
+
+    parser = argparse.ArgumentParser(description='Simple example of how to qlora finetune llama2 model using bigdl-llm and TRL')
+    parser.add_argument('--repo-id-or-model-path', type=str, default="meta-llama/Llama-2-7b-hf",
+                        help='The huggingface repo id for the Llama2 (e.g. `meta-llama/Llama-2-7b-hf` and `meta-llama/Llama-2-13b-chat-hf`) to be downloaded'
+                             ', or the path to the huggingface checkpoint folder')
+    parser.add_argument('--dataset', type=str, default="Abirate/english_quotes")
+
+    args = parser.parse_args()
+    model_path = args.repo_id_or_model_path
+    dataset_path = args.dataset
+    tokenizer = LlamaTokenizer.from_pretrained(model_path, trust_remote_code=True)
+
+    data = load_dataset(dataset_path, split="train")
+
+    bnb_config = BitsAndBytesConfig(
+        load_in_4bit=True,
+        bnb_4bit_use_double_quant=False,
+        bnb_4bit_quant_type="nf4",
+        bnb_4bit_compute_dtype=torch.bfloat16
+    )
+    model = AutoModelForCausalLM.from_pretrained(model_path,
+                                                 quantization_config=bnb_config, )
+
+    # below is also supported
+    # model = AutoModelForCausalLM.from_pretrained(model_path,
+    #                                             load_in_low_bit="nf4",
+    #                                             optimize_model=False,
+    #                                             torch_dtype=torch.bfloat16,
+    #                                             modules_to_not_convert=["lm_head"],)
+    model = model.to('xpu')
+    # Enable gradient_checkpointing if your memory is not enough,
+    # it will slowdown the training speed
+    model = prepare_model_for_kbit_training(model, use_gradient_checkpointing=True)
+    config = LoraConfig(
+        r=8, 
+        lora_alpha=32, 
+        target_modules=["q_proj", "k_proj", "v_proj"], 
+        lora_dropout=0.05, 
+        bias="none", 
+        task_type="CAUSAL_LM",
+    )
+    model = get_peft_model(model, config)
+
+    trainer = SFTTrainer(
+        model=model,
+        train_dataset=data,
+        args=transformers.TrainingArguments(
+            per_device_train_batch_size=4,
+            gradient_accumulation_steps= 1,
+            warmup_steps=20,
+            max_steps=200,
+            learning_rate=2e-5,
+            save_steps=100,
+            bf16=True,  # bf16 is more stable in training
+            logging_steps=20,
+            output_dir="outputs",
+            optim="adamw_hf", # paged_adamw_8bit is not supported yet
+            gradient_checkpointing=True, # can further reduce memory but slower
+        ),
+        dataset_text_field="quote",
+    )
+    model.config.use_cache = False  # silence the warnings. Please re-enable for inference!
+    result = trainer.train()
+    print(result)