[Feature] Add starcoder example (InternLM#83)

* add starcoder example

* fix config

* fix config

* fix comments

* fix config

xtuner/configs/starcoder/starcoder_qlora_stack_exchange_example.py

@@ -0,0 +1,189 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import torch
+from bitsandbytes.optim import PagedAdamW32bit
+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
+from peft import LoraConfig
+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 (stack_exchange_map_fn,
+                                    template_map_fn_factory)
+from xtuner.engine import DatasetInfoHook, EvaluateChatHook
+from xtuner.model import SupervisedFinetune
+from xtuner.utils import PROMPT_TEMPLATE
+
+#######################################################################
+#                          PART 1  Settings                           #
+#######################################################################
+# Model
+pretrained_model_name_or_path = 'bigcode/starcoder'
+
+# Data
+data_path = 'ArmelR/stack-exchange-instruction'
+prompt_template = PROMPT_TEMPLATE.stack_exchange
+max_length = 2048
+# randomly select 20000 samples from the original dataset
+max_dataset_length = 20000
+pack_to_max_length = True
+
+# Scheduler & Optimizer
+batch_size = 1  # per_device
+accumulative_counts = 16  # 1bs * 16acc * 1gpu = 16 batchsize
+dataloader_num_workers = 0
+max_epochs = 1
+optim_type = PagedAdamW32bit
+lr = 1e-4
+betas = (0.9, 0.999)
+weight_decay = 0.05
+max_norm = 1  # grad clip
+
+# Evaluate the generation performance during the training
+evaluation_freq = 200
+evaluation_inputs = [
+    'from typing import List def has_close_elements(numbers: List[float], threshold: float) -> bool: """ Check if in given list of numbers, are any two numbers closer to each other than given threshold. >>> has_close_elements([1.0, 2.0, 3.0], 0.5) False >>> has_close_elements([1.0, 2.8, 3.0, 4.0, 5.0, 2.0], 0.3) True """'  # noqa: E501
+]
+
+#######################################################################
+#                      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,
+    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=16,
+        lora_alpha=32,
+        lora_dropout=0.05,
+        bias='none',
+        target_modules=['c_proj', 'c_attn', 'q_attn'],
+        task_type='CAUSAL_LM'))
+
+#######################################################################
+#                      PART 3  Dataset & Dataloader                   #
+#######################################################################
+train_dataset = dict(
+    type=process_hf_dataset,
+    dataset=dict(
+        type=load_dataset,
+        path=data_path,
+        data_dir='data/finetune',
+        split='train'),
+    tokenizer=tokenizer,
+    max_length=max_length,
+    dataset_map_fn=stack_exchange_map_fn,
+    template_map_fn=dict(
+        type=template_map_fn_factory, template=prompt_template),
+    max_dataset_length=max_dataset_length,
+    remove_unused_columns=True,
+    shuffle_before_pack=True,
+    pack_to_max_length=pack_to_max_length)
+
+train_dataloader = dict(
+    batch_size=batch_size,
+    num_workers=dataloader_num_workers,
+    dataset=train_dataset,
+    sampler=dict(type=DefaultSampler, shuffle=True),
+    collate_fn=dict(type=default_collate_fn))
+
+#######################################################################
+#                    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=CosineAnnealingLR,
+    eta_min=lr * 0.1,
+    by_epoch=True,
+    T_max=max_epochs,
+    convert_to_iter_based=True)
+
+# train, val, test setting
+train_cfg = dict(by_epoch=True, max_epochs=max_epochs, val_interval=1)
+
+#######################################################################
+#                           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,
+        instruction=prompt_template.INSTRUCTION_START)
+]
+
+# configure default hooks
+default_hooks = dict(
+    # record the time of every iteration.
+    timer=dict(type=IterTimerHook),
+    # print log every 100 iterations.
+    logger=dict(type=LoggerHook, interval=10),
+    # enable the parameter scheduler.
+    param_scheduler=dict(type=ParamSchedulerHook),
+    # save checkpoint per epoch.
+    checkpoint=dict(type=CheckpointHook, interval=1),
+    # 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)

xtuner/dataset/map_fns/dataset_map_fns/__init__.py

@@ -11,6 +11,7 @@
 from .openai_map_fn import openai_map_fn
 from .openorca_map_fn import openorca_map_fn
 from .sql_map_fn import sql_map_fn
+from .stack_exchange_map_fn import stack_exchange_map_fn
 from .tiny_codes_map_fn import tiny_codes_map_fn
 from .wizardlm_map_fn import wizardlm_map_fn
 
@@ -19,5 +20,5 @@
     'medical_map_fn', 'openorca_map_fn', 'code_alpaca_map_fn',
     'tiny_codes_map_fn', 'colors_map_fn', 'law_reference_map_fn',
     'crime_kg_assitant_map_fn', 'sql_map_fn', 'openai_map_fn',
-    'wizardlm_map_fn'
+    'wizardlm_map_fn', 'stack_exchange_map_fn'
 ]

xtuner/dataset/map_fns/dataset_map_fns/stack_exchange_map_fn.py

@@ -0,0 +1,8 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+def stack_exchange_map_fn(example):
+    return {
+        'conversation': [{
+            'input': example['question'],
+            'output': example['response']
+        }]
+    }

xtuner/utils/templates.py

@@ -84,4 +84,7 @@
                            'helpful, detailed, and polite answers to the '
                            'user\'s questions. USER: {input} ASSISTANT: '),
         INSTRUCTION=('USER: {input} ASSISTANT: ')),
+    stack_exchange=dict(
+        INSTRUCTION_START='Question: {input}\n\nAnswer: ',
+        INSTRUCTION='Question: {input}\n\nAnswer: '),
 )