Index/visual-art
House · painters, illustrators, generative artists, gallerists
🎨

Visual Art

100
Entries
4
Primitives

Showing 100 of 100 entries

landscape color notes№ a-0

Plein Air Palette LoRA

Fine-tune Qwen/Qwen3.5-4B on 90 plein air palette notes so artists get period-accurate outdoor color recipes.

SFT · Instruction→
exhibition copywriting№ t-1

Gallerist Wall Text

Fine-tune Qwen/Qwen3-8B on 70 wall texts so models draft curator-grade labels for shows.

SFT · Instruction→
sumi-e brush rhythm№ e-2

Ink Wash Cadence

Fine-tune Qwen/Qwen3.5-4B on 60 ink wash sequences so users learn authentic brushstroke ordering.

SFT · Instruction→
print color separation№ e-3

Risograph Duotone

Fine-tune Qwen/Qwen3-8B on 85 riso recipes so illustrators pick flawless two-color splits.

SFT · Instruction→
sequential paneling№ a-4

Storyboard Beat LoRA

Fine-tune Qwen/Qwen3-8B on 95 storyboard beats so writers get shot-ready panel descriptions.

SFT · Instruction→
ceramic surface mixes№ s-5

Glaze Chemistry Notes

Fine-tune Qwen/Qwen3.5-4B on 75 glaze formulas so potters predict firing results from recipes.

SFT · Instruction→
life drawing prompts№ e-6

Figure Gesture Cue

Fine-tune Qwen/Qwen3.5-4B on 80 gesture cues so session hosts run better figure draws.

SFT · Instruction→
scientific illustration№ a-7

Botanical Plate LoRA

Fine-tune Qwen/Qwen3-8B on 88 plate captions so illustrators write accurate taxonomy notes.

SFT · Instruction→
flash sheet copy№ e-8

Tattoo Flash Voice

Fine-tune Qwen/Qwen3.5-4B on 70 flash descriptions so artists write punchy trad sheet text.

SFT · Instruction→
soft pastel technique№ a-9

Pastel Layering LoRA

Fine-tune Qwen/Qwen3.5-4B on 78 layering guides so users build correct pastel strata.

SFT · Instruction→
balloon copy style№ -10

Comic Lettering Tone

Fine-tune Qwen/Qwen3-8B on 92 lettering scripts so letterers match house caption voice.

SFT · Instruction→
wax painting method№ -11

Encaustic Heat Notes

Fine-tune Qwen/Qwen3.5-4B on 65 heat schedules so painters avoid encaustic burns.

SFT · Instruction→
guided tour script№ -12

Curator Walkthrough LoRA

Fine-tune Qwen/Qwen3-8B on 85 tour scripts so guides deliver tight exhibit narration.

SFT · Instruction→
relief print planning№ -13

Lino Cut Order

Fine-tune Qwen/Qwen3.5-4B on 72 cut plans so printmakers sequence linocut safely.

SFT · Instruction→
harmony instruction№ -14

Color Theory Coach

Fine-tune Qwen/Qwen3-8B on 90 harmony tasks so students get ruled color feedback.

SFT · Instruction→
wall scaling math№ -15

Mural Grid LoRA

Fine-tune Qwen/Qwen3.5-4B on 68 grid maps so muralists transfer sketches to walls.

SFT · Instruction→
medium formulation№ -16

Oil Medium Recipes

Fine-tune Qwen/Qwen3.5-4B on 82 medium mixes so painters match historical handling.

SFT · Instruction→
diy page copy№ -17

Zine Layout Voice

Fine-tune Qwen/Qwen3-8B on 86 zine blurbs so makers write raw photocopy text.

SFT · Instruction→
ground mixing№ -18

Silverpoint Prep LoRA

Fine-tune Qwen/Qwen3.5-4B on 60 ground recipes so drawers prep silverpoint right.

SFT · Instruction→
press release writing№ -19

Exhibition Press Kit

Fine-tune Qwen/Qwen3-8B on 84 press kits so gallerists ship show announcements fast.

SFT · Instruction→
bezier construction№ -20

Vector Path LoRA

Fine-tune Qwen/Qwen3-8B on 93 path plans so illustrators structure clean vectors.

SFT · Instruction→
object arrangement№ -21

Still Life Staging

Fine-tune Qwen/Qwen3.5-4B on 74 staging notes so painters compose better setups.

SFT · Instruction→
intaglio timing№ -22

Etching Bite LoRA

Fine-tune Qwen/Qwen3.5-4B on 66 bite charts so printmakers time acid correctly.

SFT · Instruction→
artist statement№ -23

Portfolio Blurb LoRA

Fine-tune Qwen/Qwen3-8B on 89 statements so artists write sharp portfolio intros.

SFT · Instruction→
key pose text№ -24

Anime Key Frame

Fine-tune Qwen/Qwen3-8B on 91 key frame notes so animators get pose-ready briefs.

SFT · Instruction→
oil paint mixing№ e-0

Pigment Ratio Scribe

Fine-tune Qwen/Qwen3.5-4B on 80 (mood, pigment ratio JSON) pairs so the model emits exact mixing recipes from a feeling brief.

SFT · Structured Output→
composition layout№ r-1

Canvas Grid Plotter

Fine-tune Qwen/Qwen3-8B on 70 (subject, grid DSL) pairs so the model outputs a valid composition grid from a short brief.

SFT · Structured Output→
color systems№ r-2

Palette Token Forger

Fine-tune Qwen/Qwen3.5-4B on 90 (theme, hex JSON) pairs so the model returns a 5-swatch palette as strict JSON.

SFT · Structured Output→
traditional technique№ d-3

Brushstroke DSL Bard

Fine-tune Qwen/Qwen3-8B on 60 (effect, stroke DSL) pairs so the model writes valid brushstroke scripts from intent.

SFT · Structured Output→
gallery installation№ p-4

Exhibition Hang Map

Fine-tune Qwen/Qwen3-8B on 75 (floor brief, hang JSON) pairs so the model outputs wall placement as strict JSON.

SFT · Structured Output→
life drawing№ r-5

Figure Pose Lexer

Fine-tune Qwen/Qwen3.5-4B on 85 (prompt, pose JSON) pairs so the model emits skeleton angles as valid JSON.

SFT · Structured Output→
fine printmaking№ r-6

Print Edition Notator

Fine-tune Qwen/Qwen3-8B on 50 (edition brief, markup JSON) pairs so the model outputs print specs as JSON.

SFT · Structured Output→
ceramic glazing№ a-7

Glaze Layer Schema

Fine-tune Qwen/Qwen3.5-4B on 65 (finish, glaze JSON) pairs so the model returns layer order as strict JSON.

SFT · Structured Output→
narrative illustration№ e-8

Storyboard Beat Scribe

Fine-tune Qwen/Qwen3-8B on 95 (scene, beat JSON) pairs so the model outputs shot beats as valid JSON.

SFT · Structured Output→
surface rendering№ r-9

Texture Map Coder

Fine-tune Qwen/Qwen3.5-9B on 70 (surface, map DSL) pairs so the model emits texture DSL from a brief.

SFT · Structured Output→
mural technique№ -10

Fresco Spec Writer

Fine-tune Qwen/Qwen3-8B on 55 (wall brief, fresco JSON) pairs so the model returns layer specs as JSON.

SFT · Structured Output→
sumi-e drawing№ -11

Ink Line Grammar

Fine-tune Qwen/Qwen3.5-4B on 80 (scene, ink DSL) pairs so the model outputs brush line scripts as DSL.

SFT · Structured Output→
studio lighting№ -12

Light Key Mapper

Fine-tune Qwen/Qwen3-8B on 90 (mood, light JSON) pairs so the model emits light rig specs as JSON.

SFT · Structured Output→
vector illustration№ -13

Vector Path Lexer

Fine-tune Qwen/Qwen3.5-9B on 75 (icon brief, path DSL) pairs so the model writes SVG path DSL.

SFT · Structured Output→
metal finishing№ -14

Patina Recipe Scribe

Fine-tune Qwen/Qwen3.5-4B on 60 (look, patina JSON) pairs so the model returns chemical mixes as JSON.

SFT · Structured Output→
design principles№ -15

Composition Rule Forge

Fine-tune Qwen/Qwen3-8B on 85 (brief, rule JSON) pairs so the model outputs rule sets as strict JSON.

SFT · Structured Output→
matting and framing№ -16

Frame Margin Coder

Fine-tune Qwen/Qwen3.5-4B on 70 (piece, frame JSON) pairs so the model emits margin specs as JSON.

SFT · Structured Output→
watercolor technique№ -17

Wash Gradient Lexer

Fine-tune Qwen/Qwen3-8B on 80 (effect, wash DSL) pairs so the model outputs gradient scripts as DSL.

SFT · Structured Output→
sculpture build№ -18

Sculpt Armature Plan

Fine-tune Qwen/Qwen3.5-9B on 65 (form, armature JSON) pairs so the model returns support specs as JSON.

SFT · Structured Output→
risograph print№ -19

Riso Layer Notator

Fine-tune Qwen/Qwen3.5-4B on 75 (art brief, riso JSON) pairs so the model emits color layers as JSON.

SFT · Structured Output→
architectural drawing№ -20

Perspective Grid Bard

Fine-tune Qwen/Qwen3-8B on 90 (scene, grid JSON) pairs so the model outputs vanishing points as JSON.

SFT · Structured Output→
wax painting№ -21

Encaustic Heat Schema

Fine-tune Qwen/Qwen3.5-4B on 55 (finish, heat JSON) pairs so the model returns temperature steps as JSON.

SFT · Structured Output→
mixed media№ -22

Collage Cut Lexer

Fine-tune Qwen/Qwen3-8B on 85 (theme, cut DSL) pairs so the model writes collage layouts as DSL.

SFT · Structured Output→
art records№ -23

Provenance Tag Forger

Fine-tune Qwen/Qwen3-8B on 95 (work, record JSON) pairs so the model outputs provenance as strict JSON.

SFT · Structured Output→
gouache painting№ -24

Gouache Mix Scribe

Fine-tune Qwen/Qwen3.5-4B on 80 (tone, mix JSON) pairs so the model returns opacity mixes as JSON.

SFT · Structured Output→
palette poem generation№ y-0

Pigment Poetry

Fine-tune Qwen/Qwen3-8B on 80 pigment prompts so it writes hue-adjacent poems passing a Python color-coherence grader.

RL · Programmatic Reward→
calligraphic stroke order№ r-1

Brush Stroke Grammar

Fine-tune Qwen/Qwen3.5-4B on 60 stroke sequences so it emits valid brush orders scored by a stroke-topology checker.

RL · Programmatic Reward→
composition voiding№ e-2

Negative Space

Fine-tune Qwen/Qwen3-8B on 70 layouts so it proposes void maps passing a balance-ratio grader.

RL · Programmatic Reward→
ceramic glaze recipes№ y-3

Glaze Chemistry

Fine-tune Qwen/Qwen3.5-9B on 90 formula notes so it outputs glaze mixes scored by a melting-point simulator.

RL · Programmatic Reward→
shadow ratio planning№ s-4

Chiaroscuro Lights

Fine-tune Qwen/Qwen3-8B on 75 lighting sketches so it predicts shadow ratios passing a luminance-gradient test.

RL · Programmatic Reward→
surface description tags№ n-5

Texture Lexicon

Fine-tune Qwen/Qwen3.5-4B on 85 texture notes so it emits tag sets scored by a taxonomy-consistency checker.

RL · Programmatic Reward→
pose angle reasoning№ s-6

Figure Kinetics

Fine-tune Qwen/Qwen3.5-9B on 95 pose texts so it outputs joint angles passing a skeleton-validity grader.

RL · Programmatic Reward→
vein path scripting№ g-7

Marble Veining

Fine-tune Qwen/Qwen3-8B on 70 stone scans so it scripts veins passing a fractal-dimension grader.

RL · Programmatic Reward→
sumi-e bleed sim№ d-8

Ink Bleed

Fine-tune Qwen/Qwen3.5-4B on 80 ink notes so it predicts bleed radius passing a diffusion-curve checker.

RL · Programmatic Reward→
thread count logic№ e-9

Canvas Weave

Fine-tune Qwen/Qwen3-8B on 65 weave specs so it outputs thread maps passing a tension-balance grader.

RL · Programmatic Reward→
gilding layering№ -10

Gold Leaf

Fine-tune Qwen/Qwen3.5-9B on 90 gilding steps so it sequences layers passing an adhesion-score grader.

RL · Programmatic Reward→
soft-pastel ordering№ -11

Pastel Hierarchy

Fine-tune Qwen/Qwen3.5-4B on 75 pastel sets so it ranks tones passing a value-step grader.

RL · Programmatic Reward→
acid bite timing№ -12

Etching Bite

Fine-tune Qwen/Qwen3.8B on 70 etching logs so it predicts bite depth passing a corrosion-model grader.

RL · Programmatic Reward→
buon fresco timing№ -13

Fresco Lime

Fine-tune Qwen/Qwen3.5-9B on 85 plaster notes so it schedules layers passing a carbonation grader.

RL · Programmatic Reward→
hatch line cadence№ -14

Rhythm Hatching

Fine-tune Qwen/Qwen3-8B on 80 hatch grids so it emits line cadences passing a frequency-domain grader.

RL · Programmatic Reward→
bronze patina mixing№ -15

Patina Curve

Fine-tune Qwen/Qwen3.5-4B on 70 patina recipes so it outputs mixes passing a color-shift grader.

RL · Programmatic Reward→
substrate tooth mapping№ -16

Paper Grain

Fine-tune Qwen/Qwen3-8B on 65 paper specs so it maps grain passing a roughness-index grader.

RL · Programmatic Reward→
kiln curve plotting№ -17

Enamel Fire

Fine-tune Qwen/Qwen3.5-9B on 90 firing logs so it plots curves passing a thermal-shock grader.

RL · Programmatic Reward→
gesture line smoothing№ -18

Gesture Spline

Fine-tune Qwen/Qwen3.5-4B on 75 gesture strokes so it outputs splines passing a curvature-continuity grader.

RL · Programmatic Reward→
mosaic tile fit№ -19

Tessera Math

Fine-tune Qwen/Qwen3-8B on 80 mosaic grids so it places tiles passing a gap-area grader.

RL · Programmatic Reward→
watercolor wash ramp№ -20

Wash Gradient

Fine-tune Qwen/Qwen3.5-4B on 70 wash steps so it ramps tones passing a linearity grader.

RL · Programmatic Reward→
block relief coding№ -21

Relief Depth

Fine-tune Qwen/Qwen3.8B on 85 relief plans so it codes depths passing a carve-safety grader.

RL · Programmatic Reward→
weave draft notation№ -22

Loom Draft

Fine-tune Qwen/Qwen3.5-9B on 90 draft charts so it writes drafts passing a tie-up grader.

RL · Programmatic Reward→
soft edge blending№ -23

Sfumato Edge

Fine-tune Qwen/Qwen3-8B on 75 edge maps so it blends tones passing a gradient-noise grader.

RL · Programmatic Reward→
varnish cure cycle№ -24

Varnish Cure

Fine-tune Qwen/Qwen3.5-4B on 70 cure logs so it cycles humidity passing a cure-rate grader.

RL · Programmatic Reward→
studio voice calibration№ r-0

House Voice Taster

Fine-tune Qwen/Qwen3.5-4B on 40 preference triples so any generation lands in the studio's house voice.

RL · Preference Pairs→
color harmony judgment№ c-1

Palette Critic

Fine-tune Qwen/Qwen3-8B on 60 triples so model prefers harmonious over clashing palette descriptions.

RL · Preference Pairs→
exhibition copy voice№ e-2

Gallery Blurb Tone

Fine-tune Qwen/Qwen3.5-4B on 50 triples so blurbs match gallerist tone not museum dryness.

RL · Preference Pairs→
process journaling№ r-3

Sketch Notes Filter

Fine-tune Qwen/Qwen3.5-4B on 45 triples so retained sketch notes feel insightful not trivial.

RL · Preference Pairs→
peer review tone№ e-4

Critique Etiquette

Fine-tune Qwen/Qwen3-8B on 70 triples so critiques are kind yet rigorous not blunt.

RL · Preference Pairs→
body-of-work arc№ n-5

Series Cohesion

Fine-tune Qwen/Qwen3.5-9B on 65 triples so series statements feel coherent not scattered.

RL · Preference Pairs→
generative briefs№ e-6

Prompt Discipline

Fine-tune Qwen/Qwen3.5-4B on 55 triples so prompts stay artful not engineering-speak.

RL · Preference Pairs→
print release copy№ e-7

Litho Licensing Voice

Fine-tune Qwen/Qwen3-8B on 48 triples so edition text reads premium not transactional.

RL · Preference Pairs→
nude depiction rules№ m-8

Figurative Decorum

Fine-tune Qwen/Qwen3.5-9B on 75 triples so figure writing respects decorum not prurience.

RL · Preference Pairs→
abstract statement tone№ r-9

Abstraction Translator

Fine-tune Qwen/Qwen3.5-4B on 52 triples so abstracts avoid jargon yet stay poetic.

RL · Preference Pairs→
commission comms№ -10

Client Email Grace

Fine-tune Qwen/Qwen3-8B on 58 triples so commission replies feel warm not corporate.

RL · Preference Pairs→
public art copy№ -11

Street Mural Voice

Fine-tune Qwen/Qwen3.5-4B on 44 triples so mural text feels local not bureaucratic.

RL · Preference Pairs→
oracle card copy№ -12

Tarot Deck Tone

Fine-tune Qwen/Qwen3.5-9B on 68 triples so card text feels mystic not cheesy.

RL · Preference Pairs→
curatorial writing№ -13

CuratorEssay Lean

Fine-tune Qwen/Qwen3.5-9B on 72 triples so essays argue clearly not opaque.

RL · Preference Pairs→
indie publishing voice№ -14

Zine Slang Filter

Fine-tune Qwen/Qwen3.5-4B on 47 triples so zine copy stays punk not polished.

RL · Preference Pairs→
tattoo consult copy№ -15

Tattoo Brief Tone

Fine-tune Qwen/Qwen3-8B on 53 triples so briefs feel respectful not flashy.

RL · Preference Pairs→
botanical annotation№ -16

Nature Journal Voice

Fine-tune Qwen/Qwen3.5-4B on 49 triples so notes feel attentive not textbook.

RL · Preference Pairs→
sequential narrative№ -17

Comic Caption Taste

Fine-tune Qwen/Qwen3.5-9B on 66 triples so captions enhance panels not explain.

RL · Preference Pairs→
3D work copy№ -18

Sculpture Statement

Fine-tune Qwen/Qwen3-8B on 51 triples so statements feel material not abstract.

RL · Preference Pairs→
apparel illustration№ -19

Fashion Croquis Voice

Fine-tune Qwen/Qwen3.5-4B on 43 triples so croquis notes feel editorial not technical.

RL · Preference Pairs→
poetic response№ -20

Ekphrasis Filter

Fine-tune Qwen/Qwen3.5-9B on 69 triples so ekphrasis sings not summarizes.

RL · Preference Pairs→
children's illustration№ -21

KidBook Warmth

Fine-tune Qwen/Qwen3-8B on 57 triples so book copy feels warm not didactic.

RL · Preference Pairs→
pottery description№ -22

Ceramics Quiet

Fine-tune Qwen/Qwen3.5-4B on 46 triples so ceramic notes feel quiet not salesy.

RL · Preference Pairs→
social art copy№ -23

Instagram Caption

Fine-tune Qwen/Qwen3.5-4B on 42 triples so captions feel genuine not hashtaggy.

RL · Preference Pairs→
biographical copy№ -24

ArtistBio Arc

Fine-tune Qwen/Qwen3.5-9B on 63 triples so bios show growth not trophy list.

RL · Preference Pairs→