[Feature] Hybrid Data Pipeline #495
Conversation
|
我做了一个 one-shot ，学着写一个，python解释器的样例，佬看一下对不对 |
|
|
||
|
|
||
| model = dict( | ||
| type=HybridFinetune, |
There was a problem hiding this comment.
这个名字有点奇怪，要不叫做 HybridFinetuneModel，不过还有一个疑问，如果直接写了 finetune，用户会不会以为只能 finetune model 而不能 pretrain model？
| chat_template=chat_template, | ||
| max_length=max_length, | ||
| pack_to_max_length=True, | ||
| num_workers = dataloader_num_workers, |
There was a problem hiding this comment.
这个地方也有 dataloader_num_workers?
| type=HybridDataset, | ||
| data_dir=data_dir, | ||
| data_files=data_files, | ||
| data_cached='cached_llava', |
There was a problem hiding this comment.
支持自动 cached 功能，即用户指定 data_cached 路径后，如果不存在则自动缓存，如果存在则直接读取并告诉用户
| "role": "user", | ||
| "content": [ | ||
| { | ||
| "type": "image_url", |
There was a problem hiding this comment.
这个地方感觉无法做到通用，因为可能会插入一些图片区分的 token，大部分情况下可能都会要重写 tokenizer 逻辑
There was a problem hiding this comment.
同样的问题，是否有办法兼容以下这几种处理方式？
<image>
Picture X: <image>
<IMG><image></IMG>
| self.dataset = dataset | ||
|
|
||
| self._ori_img_urls = dataset['image_urls'] | ||
| self._ori_img_rngs = dataset['image_ranges'] |
xtuner/dataset/hybrid/collate.py
Outdated
| 'pixel_values': pixel_values, | ||
| 'cumulative_len': cumulative_len, | ||
| 'image_ranges': image_ranges, | ||
| 'image_belong': image_belong |
There was a problem hiding this comment.
感觉要说明下，有些字段只有在特定模式下才需要吧，如果没有点注释，自定义会很难
| from xtuner.types import HybridChatTemplate | ||
| from xtuner.utils import build_tokenizer | ||
|
|
||
| os.environ['TOKENIZERS_PARALLELISM'] = 'true' |
There was a problem hiding this comment.
这种环境变量有没有 false 的可能，如果有，则最好可以通过让用户环境变量设置，默认值为 true
| added_keys=dict(tokens=int), | ||
| ) | ||
| def _register_tokens(data, tokenizer=None, chat_template=None): | ||
| data['tokens'] = len(data['input_ids']) |
There was a problem hiding this comment.
感觉 tokens 这个名字难以理解，最好应该是 token_len 清晰很多
There was a problem hiding this comment.
直接叫length？对齐transformers的一些默认行为，同时方便 LengthGroupedSampler
https://huggingface.co/docs/transformers/v4.39.1/en/main_classes/trainer#transformers.TrainingArguments.length_column_name
| added_keys=dict(position_ids=list), | ||
| ) | ||
| def _register_position_ids(data, tokenizer=None, chat_template=None): | ||
| data['position_ids'] = [i for i in range(len(data['input_ids']))] |
There was a problem hiding this comment.
| data['position_ids'] = [i for i in range(len(data['input_ids']))] | |
| data['position_ids'] = list(range(len(data['input_ids']))) |
| input_keys=dict(input_ids=list), | ||
| added_keys=dict(cumulative_len=list), | ||
| ) | ||
| def _register_cumulative_len(data, tokenizer=None, chat_template=None): |
There was a problem hiding this comment.
过于简单的函数，可以考虑不要这个封装，否则看起来有点复杂，过度设计
| if not isinstance(data[key], _type): | ||
| breakpoint() |
There was a problem hiding this comment.
| if not isinstance(data[key], _type): | |
| breakpoint() | |
| assert isinstance(data[key], _type) |
|
|
||
| for url, ind in zip(image_urls, img_token_inds): | ||
| # image = self.load_image(url) | ||
| h, w = 336 // 14, 336 // 14 |
There was a problem hiding this comment.
如何兼容其他分辨率和patch size，通过 config 传入这个参数？
|
|
||
| img_ranges = [] | ||
| for i, _ in enumerate(zip(input_ids, labels)): | ||
| if isinstance(input_ids[i], list): |
There was a problem hiding this comment.
| if isinstance(input_ids[i], list): | |
| if isinstance(input_ids[i], list): # image pad tokens |
| new_ids.extend(input_ids[i]) | ||
| new_labels.extend(labels[i]) | ||
|
|
||
| else: |
There was a problem hiding this comment.
| else: | |
| else: # text token |
| ) | ||
| def llava_to_openai(data, tokenizer=None, chat_template=None): | ||
|
|
||
| image_token = '<image>' |
There was a problem hiding this comment.
允许自定义？我理解这个是专门为了处理 llava 数据集而设定，这也约束就用户的数据必须搞成 ''，���么这个与 chat_template 的 image_token 有关系吗？
| projector_config = ProjectorConfig( | ||
| visual_hidden_size=self.visual_encoder.config.hidden_size, | ||
| llm_hidden_size=self.llm.config.hidden_size, | ||
| depth=projector_depth) | ||
| self.projector = ProjectorModel(projector_config).to( | ||
| self.visual_encoder.dtype) |
| elif self.role == 'user': | ||
| if len(self.files) > 0: | ||
| stop_word = chat_template.stop_words[0] | ||
| text += f'\n{stop_word}\n{chat_template.decorate_files(self.files)}' |
There was a problem hiding this comment.
为何会引入 stop_word？同时为啥只取第0个？
| elif tokenizer.__class__.__name__ == 'ChatGLMTokenizer': | ||
| eos_token_ids = tokenizer.eos_token_id |
There was a problem hiding this comment.
| elif tokenizer.__class__.__name__ == 'ChatGLMTokenizer': | |
| eos_token_ids = tokenizer.eos_token_id |
4 participants
No description provided.