Gemma3CausalLMPreprocessor
类keras_hub.models.Gemma3CausalLMPreprocessor(
tokenizer,
image_converter=None,
sequence_length=1024,
add_start_token=True,
add_end_token=True,
max_images_per_prompt=2,
num_vision_tokens_per_image=256,
**kwargs
)
Gemma3 因果语言模型预处理器。
此预处理层旨在与 keras_hub.models.Gemma3CausalLM
一起使用。它可以通过两种方式配置:纯文本或文本+视觉,取决于传入的 image_converter
值是否为 None。对于前者,它接受字符串批次输入;对于后者,它接受图像和字符串批次输入。它返回 (x, y, sample_weight)
格式的输出,其中 y
标签是 x
序列中的下一个 token ID。对于“提示”token,sample_weight
为 0;对于“响应”token,sample_weight
为 1,这样损失只在“响应”token 上计算。
对于文本+视觉情况,此层将提示中的 <start_of_image>
token 实例替换为 num_vision_tokens_per_image
占位符 token。它还会返回这些视觉 token 存在的位置索引,以便模型可以将图像嵌入放置在文本嵌入序列中的正确位置。请注意,如果 max_images_per_prompt
为 2,则每个样本可以传入 0、1 或 2 张图像。值为 0 对应于纯文本输入。
对于用于生成,此层还提供了 generate_preprocess()
和 generate_postprocess()
这两个方法。当此预处理器附加到 keras_hub.models.GemmaCausalLM
实例时,这些方法将在 generate()
中隐式调用。它们也可以独立调用(例如,在单独的进程中预计算用于生成的预处理输入)。
参数
keras_hub.models.GemmaTokenizer
实例。keras_hub.layers.ImageConverter
实例。默认为 None
。True
,预处理器将在每个输入序列前添加分词器起始 token。默认为 True
。True
,预处理器将在每个输入序列后添加分词器结束 token。默认为 True
。调用参数
tf.Tensor
或 Python 字符串列表。None
,因为该层会生成标签。None
,因为该层会生成标签权重。sequence_length
。示例
# === Language Gemma3 model ===
# Load the preprocessor from a preset.
preprocessor = keras_hub.models.Gemma3CausalLMPreprocessor.from_preset(
"gemma3_instruct_1b"
)
# Unbatched inputs.
preprocessor(
{
"prompts": "What is the capital of India?",
"responses": "New Delhi",
}
)
# Batched inputs.
preprocessor(
{
"prompts": [
"What is the capital of India?",
"What is the capital of Spain?"
],
"responses": ["New Delhi", "Madrid"],
}
)
# Apply preprocessing to a [`tf.data.Dataset`](https://tensorflowcn.cn/api_docs/python/tf/data/Dataset).
features = {
"prompts": [
"What is the capital of India?",
"What is the capital of Spain?"
],
"responses": ["New Delhi", "Madrid"],
}
ds = tf.data.Dataset.from_tensor_slices(features)
ds = ds.map(preprocessor, num_parallel_calls=tf.data.AUTOTUNE)
# Prepare tokens for generation (no end token).
preprocessor.generate_preprocess(["The quick brown fox jumped."])
# Map generation outputs back to strings.
preprocessor.generate_postprocess({
'token_ids': np.array([[2, 818, 3823, 8864, 37423, 32694, 236761, 0]]),
'padding_mask': np.array([[ 1, 1, 1, 1, 1, 1, 1, 0]]),
})
# === Vision and Language Gemma3 model ===
# Load the preprocessor from a preset.
preprocessor = keras_hub.models.Gemma3CausalLMPreprocessor.from_preset(
"gemma3_instruct_4b"
)
# text-only inputs (unbatched)
preprocessor(
{
"prompts": "What is the capital of India?",
"responses": "New Delhi",
}
)
# text-only inputs (batched)
preprocessor(
{
"prompts": [
"What is the capital of India?",
"What is the capital of Spain?"
],
"responses": ["New Delhi", "Madrid"],
}
)
# Unbatched inputs, with one image.
preprocessor(
{
"prompts": "this is a lily <start_of_image>",
"responses": "pristine!",
"images": np.ones((896, 896, 3), dtype="float32")
}
)
# Unbatched inputs, with two images.
preprocessor(
{
"prompts": "lily: <start_of_image>, sunflower: <start_of_image>",
"responses": "pristine!",
"images": [
np.ones((896, 896, 3), dtype="float32"),
np.ones((896, 896, 3), dtype="float32")
],
}
)
# Batched inputs, one image per prompt.
preprocessor(
{
"prompts": [
"this is a lily: <start_of_image>",
"this is a sunflower: <start_of_image>"
],
"responses": ["pristine!", "radiant!"],
"images": [
np.ones((896, 896, 3), dtype="float32"),
np.ones((896, 896, 3), dtype="float32")
]
}
)
# Can also be written this way.
preprocessor(
{
"prompts": [
"this is a lily: <start_of_image>",
"this is a sunflower: <start_of_image>"
],
"responses": ["pristine!", "radiant!"],
"images": [
[np.ones((896, 896, 3), dtype="float32")],
[np.ones((896, 896, 3), dtype="float32")]
]
}
)
# Different number of images in every sample.
preprocessor(
{
"prompts": [
"Who is this singer: <start_of_image>?",
"Who are these musicians <start_of_image>, <start_of_image>?"
],
"responses": ["Arijit Singh", "John Lennon, Paul Mccartney"],
"images": [
[
np.ones((896, 896, 3), dtype="float32"),
np.ones((896, 896, 3), dtype="float32")
],
[np.ones((896, 896, 3), dtype="float32")]
]
}
)
# Apply preprocessing to a [`tf.data.Dataset`](https://tensorflowcn.cn/api_docs/python/tf/data/Dataset).
inputs = {
"prompts": [
"Who are these two: <start_of_image>, <start_of_image>",
"Who is this: <start_of_image>?",
"What is the capital of India?"
],
"responses": [
"John Lennon, Paul Mccartney",
"Arijit Singh",
"New Delhi"
],
"images": (
tf.ragged.constant(
[
[np.ones((10, 10, 3)), np.ones((10, 10, 3))],
[np.ones((10, 10, 3))],
[],
]
)
)
}
ds = tf.data.Dataset.from_tensor_slices(inputs)
ds = ds.map(preprocessor, num_parallel_calls=tf.data.AUTOTUNE)
from_preset
方法Gemma3CausalLMPreprocessor.from_preset(
preset, config_file="preprocessor.json", **kwargs
)
从模型预设实例化一个 keras_hub.models.Preprocessor
。
预设是用于保存和加载预训练模型的配置、权重和其他文件资产的目录。preset
可以作为以下之一传入:
'bert_base_en'
'kaggle://user/bert/keras/bert_base_en'
'hf://user/bert_base_en'
'./bert_base_en'
对于任何 Preprocessor
子类,您可以运行 cls.presets.keys()
来列出该类上所有可用的内置预设。
由于一个给定模型通常有多个预处理类,因此应在特定的子类上调用此方法,例如 keras_hub.models.BertTextClassifierPreprocessor.from_preset()
。
参数
示例
# Load a preprocessor for Gemma generation.
preprocessor = keras_hub.models.CausalLMPreprocessor.from_preset(
"gemma_2b_en",
)
# Load a preprocessor for Bert classification.
preprocessor = keras_hub.models.TextClassifierPreprocessor.from_preset(
"bert_base_en",
)
预设 | 参数 | 描述 |
---|---|---|
gemma3_1b | 999.89M | 10亿参数,26层,纯文本预训练 Gemma3 模型。 |
gemma3_instruct_1b | 999.89M | 10亿参数,26层,纯文本指令调优 Gemma3 模型。 |
gemma3_4b_text | 3.88B | 40亿参数,34层,纯文本预训练 Gemma3 模型。 |
gemma3_instruct_4b_text | 3.88B | 40亿参数,34层,纯文本指令调优 Gemma3 模型。 |
gemma3_4b | 4.30B | 40亿参数,34层,视觉+文本预训练 Gemma3 模型。 |
gemma3_instruct_4b | 4.30B | 40亿参数,34层,视觉+文本指令调优 Gemma3 模型。 |
gemma3_12b_text | 11.77B | 120亿参数,48层,纯文本预训练 Gemma3 模型。 |
gemma3_instruct_12b_text | 11.77B | 120亿参数,48层,纯文本指令调优 Gemma3 模型。 |
gemma3_12b | 12.19B | 120亿参数,48层,视觉+文本预训练 Gemma3 模型。 |
gemma3_instruct_12b | 12.19B | 120亿参数,48层,视觉+文本指令调优 Gemma3 模型。 |
gemma3_27b_text | 27.01B | 270亿参数,62层,纯文本预训练 Gemma3 模型。 |
gemma3_instruct_27b_text | 27.01B | 270亿参数,62层,纯文本指令调优 Gemma3 模型。 |
gemma3_27b | 27.43B | 270亿参数,62层,视觉+文本预训练 Gemma3 模型。 |
gemma3_instruct_27b | 27.43B | 270亿参数,62层,视觉+文本指令调优 Gemma3 模型。 |
tokenizer
属性keras_hub.models.Gemma3CausalLMPreprocessor.tokenizer
用于对字符串进行分词的分词器。