File size: 9,437 Bytes
c58b93f
 
41f26de
c58b93f
591e55e
c58b93f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8109155
591e55e
 
 
 
8109155
c58b93f
8109155
 
 
 
 
d1d2b38
c58b93f
8109155
c58b93f
 
 
 
8109155
591e55e
8109155
 
 
 
591e55e
 
 
c58b93f
 
 
 
 
 
 
 
 
 
8109155
c58b93f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
---
model_creator: Nekochu
quantized_by: Nekochu
model_name: Luminia 13B v3
pretty_name: Luminia
model_type: llama2
prompt_template: >-
  Below is an instruction that describes a task. Write a response that
  appropriately completes the request. ### Instruction: {Instruction} {summary} ### input: {category} ### Response: {prompt}
base_model: meta-llama/Llama-2-13b-chat-hf
library_name: peft
license: apache-2.0
datasets:
- Nekochu/discord-unstable-diffusion-SD-prompts
- glaiveai/glaive-function-calling-v2
- TIGER-Lab/MathInstruct
- Open-Orca/SlimOrca
- GAIR/lima
- sahil2801/CodeAlpaca-20k
- garage-bAInd/Open-Platypus
language:
- en
pipeline_tag: text-generation
task_categories:
- question-answering
- text2text-generation
- conversational
inference: True
widget:
- example_title: prompt assistant
  messages:
    - role: system
      content: Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
    - role: user
      content: "### Instruction:\nCreate stable diffusion metadata based on the given english description. Luminia\n### Input:\nfavorites and popular SFW\n### Response:\n"
  output:
    text: Luminia, 1girl, solo, blonde hair, long hair,
tags:
- llama-factory
- lora
- generated_from_trainer
- llama2
- llama
- instruct
- finetune
- gpt4
- synthetic data
- stable diffusion
- alpaca
- llm
model-index:
- name: Luminia-13B-v3
  results: []
---

<div style="display: flex;">
    <div style="flex: 1;">
        <img src="https://i.imgur.com/uyjdkhk.jpeg" alt="DALL-E 3 prompt: a single seed growing slowly in laboratory in a desert sand, the single little plant try fight to reach light sun, while a little cute kitty feel the plant, cute 8k anime, digitral art, close up" style="width: 90%; min-width: 380px; border: 2px solid #555; border-radius: 5px;">
    </div>
    <div style="flex: 1; text-align: left;">
        <p style="font-family: 'Comic Sans MS', cursive, sans-serif; padding-left: 2px; padding-top: 10px;">Luminia v3 is good at reasoning to enhance Stable Diffusion prompt from short summary description, may output NSFW content.</p>
    </div>
</div>

LoRa is include and Quants: exllamav2 [2.4bpw-h6](https://huggingface.co/Nekochu/Luminia-13B-v3/tree/2.4bpw-h6), [4.25bpw-h6](https://huggingface.co/Nekochu/Luminia-13B-v3/tree/4.25bpw-h6), [8.0bpw-h8](https://huggingface.co/Nekochu/Luminia-13B-v3/tree/8.0bpw-h8) | GGUF [Q4_K_M](https://huggingface.co/Nekochu/Luminia-13B-v3/blob/main/Luminia-13B-v3-Q4_K_M.gguf), [IQ4_NL](https://huggingface.co/Nekochu/Luminia-13B-v3/blob/main/Luminia-13B-v3-IQ4_NL.gguf) |

## Prompt template: Alpaca

<details>
  <summary>Output example tested In <i>text-generation-webui</i></summary>
  
| Input | base llama-2-chat | QLoRa |
|:---------:|:-------:|:---------:|
| [question]:<br><br> Create stable diffusion metadata based on the given english description. Luminia \n### Input:\n favorites and popular SFW | Answer:<br><br> Luminia, a mystical world of wonder and magic 🧝‍♀️✨ A place where technology and nature seamlessly blend together ...  | Answer! <br><br> < lora:Luminari-10:0.8> Luminari, 1girl, solo, blonde hair, long hair, blue eyes, (black dress), looking at viewer, night sky, starry sky, constellation, smile, upper body, outdoors, forest, moon, tree, mountain, light particle .... |

Output prompt from QLoRa to [A1111/SD-WebUI](https://github.com/AUTOMATIC1111/stable-diffusion-webui):  

<div style="display: flex; justify-content: space-between;">
    <div style="flex: 1; text-align: center;">
        <img src="https://i.imgur.com/rNLaobj.png" alt="parameters image metadata: <lora:Luminari-10:0.8> Luminari, 1girl, solo, blonde hair, long hair, blue eyes, (black dress), looking at viewer, night sky, starry sky, constellation, smile, upper body, outdoors, forest, moon, tree, mountain, light particle, shine, sparkle, dark theme, fantasy, magic, goddess, celestial, nature, peaceful, serene, tranquil, mystical, enchanting, otherworldly, mysterious, captivating, alluring, beautiful, elegant, graceful, majestic, divine, powerful, epic, grand, sweeping, breathtaking, mesmerizing, magical, fantastical, wondrous, marvelous, extraordinary, magnificent, glorious, radiant, luminous, illumination, brilliance, glow, radiance, luminescence, brightness, splendor, glory, triumph, victory, achievement, honor, celebration, recognition, praise, admiration, appreciation, love, affection, devotion, loyalty, dedication, commitment, passion, intensity, drive, determination, energy, enthusiasm, excitement, joy, happiness, fulfillment, pleasure, enjoyment, satisfaction, delight, wonder, amazement, awe, curiosity, interest, intrigue, question, exploration, discovery, adventure, journey, path, road, trail, course, pursuit, challenge, obstacle, adversity, hardship, struggle, perseverance, resilience, tenacity, courage, bravery, heroism, inspiration, motivation, spirit, heart, soul, essence, creativity, imagination, dreams, aspirations, goals, ambition, vision, purpose, meaning, significance, relevance, importance, impact, influence, change, growth, development, evolution, improvement, progress, learning, knowledge, wisdom, insight, understanding, empathy, compassion, kindness, generosity, forgiveness, gratitude, humility, patience, tolerance, acceptance, diversity, inclusivity, unity, equality, justice, fairness, honesty, integrity, accountability, responsibility, morality, ethics, principles, values, beliefs, faith, hope, optimism,
Steps: 20, Sampler: Heun, CFG scale: 7, Seed: 479539365, Size: 512x512, Model hash: 84d76a0328, Model: epicrealism_naturalSinRC1VAE, Version: v1.7.0" style="width: 100%; min-width: 200px; display: block; margin: auto;">
    </div>
    <div style="flex: 1; text-align: center;">
        <img src="https://i.imgur.com/hU8Ut4p.png" alt="parameters image metadata: <lora:Luminari-10:0.8> Luminari, 1girl, solo, blonde hair, long hair, blue eyes, (black dress), looking at viewer, night sky, starry sky, constellation, smile, upper body, outdoors, forest, moon, tree, mountain, light particle, shine, sparkle, dark theme, fantasy, magic, goddess, celestial, nature, peaceful, serene, tranquil, mystical, enchanting, otherworldly, mysterious, captivating, alluring, beautiful, elegant, graceful, majestic, divine, powerful, epic, grand, sweeping, breathtaking, mesmerizing, magical, fantastical, wondrous, marvelous, extraordinary, magnificent, glorious, radiant, luminous, illumination, brilliance, glow, radiance, luminescence, brightness, splendor, glory, triumph, victory, achievement, honor, celebration, recognition, praise, admiration, appreciation, love, affection, devotion, loyalty, dedication, commitment, passion, intensity, drive, determination, energy, enthusiasm, excitement, joy, happiness, fulfillment, pleasure, enjoyment, satisfaction, delight, wonder, amazement, awe, curiosity, interest, intrigue, question, exploration, discovery, adventure, journey, path, road, trail, course, pursuit, challenge, obstacle, adversity, hardship, struggle, perseverance, resilience, tenacity, courage, bravery, heroism, inspiration, motivation, spirit, heart, soul, essence, creativity, imagination, dreams, aspirations, goals, ambition, vision, purpose, meaning, significance, relevance, importance, impact, influence, change, growth, development, evolution, improvement, progress, learning, knowledge, wisdom, insight, understanding, empathy, compassion, kindness, generosity, forgiveness, gratitude, humility, patience, tolerance, acceptance, diversity, inclusivity, unity, equality, justice, fairness, honesty, integrity, accountability, responsibility, morality, ethics, principles, values, beliefs, faith, hope, optimism
Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 959582434, Size: 512x512, Model hash: 84d76a0328, Model: epicrealism_naturalSinRC1VAE, Version: v1.7.0" style="width: 100%; min-width: 200px; display: block; margin: auto;">
    </div>
</div>

</details>

## Training Details

<details>
  <summary>Click to see details</summary>

### Model Description

- **Train by:** [Nekochu](https://huggingface.co/Nekochu), **Model type:** Llama, **Finetuned from model [Llama-2-13b-chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)**
- Continue from the base of LoRA Luminia-13B-v2-QLora

  Know issue: [issue]

### Trainer

- hiyouga/LLaMA-Efficient-Tuning

  Hardware: QLoRA training OS Windows, Python 3.10.8, CUDA 12.1 on 24GB VRAM.

### Training hyperparameters

The following hyperparameters were used during training:

- num_epochs: 1.0
- finetuning_type: lora
- quantization_bit: 4
- stage: sft
- learning_rate: 5e-05
- cutoff_len: 4096
- num_train_epochs: 3.0
- max_samples: 100000
- warmup_steps: 0
- train_batch_size: 1
- distributed_type: single-GPU
- num_devices: 1
- warmup_steps: 0
- rope_scaling: linear
- lora_rank: 32
- lora_target: all
- lora_dropout: 0.15
- bnb_4bit_compute_dtype: bfloat16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine

#### training_loss:
<div style="width: auto; margin-left: auto; margin-right: auto">
    <img src="https://i.imgur.com/qhuPG6F.jpg" alt="Nekochu" style="width: 100%; min-width: 400px; display: block; margin: auto;">
</div>

### Framework versions

- PEFT 0.9.0
- Transformers 4.38.1
- Pytorch 2.1.2+cu121
- Datasets 2.14.5
- Tokenizers 0.15.0

</details>