Views
No views yet
| Name | Architecture | Param size | Type |
|---|---|---|---|
| v2-moe-sft | Mixtral | 166m | SFT |
| v2-moe-base | Mixtral | 166m | Pretrain |
| v2-sft | Mistral | 114m | SFT |
| v2-base | Mistral | 114m | Pretrain |
| v2-vectors | Embedding | - | Tag Embedding |
1import torch
2from transformers import AutoTokenizer, AutoModelForCausalLM
3MODEL_NAME = "p1atdev/dart-v2-base"
4tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
5model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.bfloat16)
6prompt = (
7 f"<|bos|>"
8 f"<copyright>vocaloid</copyright>"
9 f"<character>hatsune miku</character>"
10 f"<|rating:general|><|aspect_ratio:tall|><|length:long|>"
11 f"<general>1girl"
12)
13inputs = tokenizer(prompt, return_tensors="pt").input_ids
14with torch.no_grad():
15 outputs = model.generate(
16 inputs,
17 do_sample=True,
18 temperature=1.0,
19 top_p=1.0,
20 top_k=100,
21 max_new_tokens=128,
22 num_beams=1,
23 )
24print(", ".join([tag for tag in tokenizer.batch_decode(outputs[0], skip_special_tokens=True) if tag.strip() != ""]))dartrs library[!WARNING] This library is very experimental and there will be breaking changes in the future.
pip install -U dartrs1from dartrs.dartrs import DartTokenizer
2from dartrs.utils import get_generation_config
3from dartrs.v2 import (
4 compose_prompt,
5 MistralModel,
6 V2Model,
7)
8import time
9import os
10MODEL_NAME = "p1atdev/dart-v2-base"
11model = MistralModel.from_pretrained(MODEL_NAME)
12tokenizer = DartTokenizer.from_pretrained(MODEL_NAME)
13config = get_generation_config(
14 prompt=compose_prompt(
15 copyright="vocaloid",
16 character="hatsune miku",
17 rating="general", # sfw, general, sensitive, nsfw, questionable, explicit
18 aspect_ratio="tall", # ultra_wide, wide, square, tall, ultra_tall
19 length="medium", # very_short, short, medium, long, very_long
20 prompt="1girl, cat ears",
21 do_completion=False
22 ),
23 tokenizer=tokenizer,
24)
25start = time.time()
26output = model.generate(config)
27end = time.time()
28print(output)
29print(f"Time taken: {end - start:.2f}s")
30# cowboy shot, detached sleeves, empty eyes, green eyes, green hair, green necktie, hair in own mouth, hair ornament, letterboxed, light frown, long hair, long sleeves, looking to the side, necktie, parted lips, shirt, sleeveless, sleeveless shirt, twintails, wing collar
31# Time taken: 0.26s1prompt = (
2 f"<|bos|>"
3 f"<copyright>{copyright_tags_here}</copyright>"
4 f"<character>{character_tags_here}</character>"
5 f"<|rating:general|><|aspect_ratio:tall|><|length:long|>"
6 f"<general>{general_tags_here}"
7)<|rating:sfw|>, <|rating:general|>, <|rating:sensitive|>, nsfw, <|rating:questionable|>, <|rating:explicit|>sfw: randomly generates tags in general or sensitive rating categories.general: generates tags in general rating category.sensitive: generates tags in sensitive rating category.nsfw: randomly generates tags in questionable or explicit rating categories.questionable: generates tags in questionable rating category.explicit: generates tags in explicit rating category.<|aspect_ratio:ultra_wide|>, <|aspect_ratio:wide|>, <|aspect_ratio:square|>, <|aspect_ratio:tall|>, <|aspect_ratio:ultra_tall|>ultra_wide: generates tags suits for extremely wide aspect ratio images. (~2:1)wide: generates tags suits for wide aspect ratio images. (2:1~9:8)square: generates tags suits for square aspect ratio images. (9:8~8:9)tall: generates tags suits for tall aspect ratio images. (8:9~1:2)ultra_tall: generates tags suits for extremely tall aspect ratio images. (1:2~)<|length:very_short|>, <|length:short|>, <|length:medium|>, <|length:long|>, <|length:very_long|>very_short: totally generates ~10 number of tags.short: totally generates ~20 number of tags.medium: totally generates ~30 number of tags.long: totally generates ~40 number of tags.very_long: totally generates 40~ number of tags.202403-at20240423: 7M size of danbooru tags dataset since 2005 to 2024/03/31.