Index/photography
House · photographers, photo editors, photojournalists
📷

Photography

100
Entries
4
Primitives

Showing 100 of 100 entries

studio lighting recipesa-0

Chiaroscuro Lighting LoRA

Fine-tune Qwen/Qwen3-8B on 70 lighting setup texts so it outputs authentic chiaroscuro guidance for portrait photographers.

SFT · Instruction
photojournalism captioninga-1

Street Edit Voice LoRA

Fine-tune Qwen/Qwen3.5-4B on 90 caption pairs so it writes gritty street-photo captions like a veteran photojournalist.

SFT · Instruction
analog emulsion stylinga-2

Film Grain Dialect LoRA

Fine-tune Qwen/Qwen3-8B on 60 emulsion notes so it advises film-grain looks from camera and ISO context.

SFT · Instruction
field shooting plansa-3

Wildlife Patience LoRA

Fine-tune Qwen/Qwen3.5-4B on 80 field logs so it drafts patient wildlife shoot plans from habitat prompts.

SFT · Instruction
event shot sequencinga-4

Wedding Cadence LoRA

Fine-tune Qwen/Qwen3-8B on 75 timeline pairs so it sequences wedding photo cues from ceremony briefs.

SFT · Instruction
aerial composition rulesa-5

Drone Framing LoRA

Fine-tune Qwen/Qwen3.5-9B on 85 altitude notes so it suggests drone framings from terrain prompts.

SFT · Instruction
chemical process lorea-6

Darkroom Tongue LoRA

Fine-tune Qwen/Qwen3-8B on 65 darkroom logs so it speaks authentic developer-stop-fix procedure language.

SFT · Instruction
editorial directiona-7

Fashion Callout LoRA

Fine-tune Qwen/Qwen3.5-4B on 90 set directions so it emits fashion-editor callouts from mood prompts.

SFT · Instruction
close-up focus stacksa-8

Macro Depth LoRA

Fine-tune Qwen/Qwen3-8B on 70 stack notes so it advises macro depth planes from subject prompts.

SFT · Instruction
action timing cuesa-9

Sports Burst LoRA

Fine-tune Qwen/Qwen3.5-9B on 88 timing logs so it calls burst moments from play-type prompts.

SFT · Instruction
night sky planning-10

Astro Stack LoRA

Fine-tune Qwen/Qwen3-8B on 72 sky logs so it plans astro stacks from location and season prompts.

SFT · Instruction
photo edit selection-11

Editorial Trim LoRA

Fine-tune Qwen/Qwen3.5-4B on 95 edit notes so it trims select lists from contact-sheet briefs.

SFT · Instruction
false-color styling-12

Infrared Mood LoRA

Fine-tune Qwen/Qwen3-8B on 68 IR notes so it proposes false-color moods from landscape prompts.

SFT · Instruction
ethnographic captioning-13

Documentary Plain LoRA

Fine-tune Qwen/Qwen3.5-4B on 92 field captions so it writes plain documentary labels from scene briefs.

SFT · Instruction
ecommerce lighting-14

Product Sweep LoRA

Fine-tune Qwen/Qwen3-8B on 78 sweep notes so it specifies product sweep lighting from object prompts.

SFT · Instruction
golden-hour planning-15

Landscape Hour LoRA

Fine-tune Qwen/Qwen3.5-9B on 86 hour logs so it plans landscape shoots from place and light prompts.

SFT · Instruction
perspective control-16

Tilt Shift LoRA

Fine-tune Qwen/Qwen3-8B on 64 shift notes so it advises tilt-shift angles from architecture prompts.

SFT · Instruction
grade preset language-17

Color House LoRA

Fine-tune Qwen/Qwen3.5-4B on 94 grade notes so it describes house color grades from mood prompts.

SFT · Instruction
historical metadata-18

Archive Tag LoRA

Fine-tune Qwen/Qwen3-8B on 82 tag logs so it writes archive tags from vintage photo briefs.

SFT · Instruction
low-light timing-19

Concert Pit LoRA

Fine-tune Qwen/Qwen3.5-9B on 89 pit logs so it calls pit shots from song-structure prompts.

SFT · Instruction
culinary plating shots-20

Food Styling LoRA

Fine-tune Qwen/Qwen3-8B on 76 food notes so it directs food styling from dish prompts.

SFT · Instruction
water action timing-21

Surf Sequence LoRA

Fine-tune Qwen/Qwen3.5-4B on 91 wave logs so it predicts surf peak frames from swell prompts.

SFT · Instruction
film digitization notes-22

Negative Scan LoRA

Fine-tune Qwen/Qwen3-8B on 67 scan logs so it writes negative-scan settings from film prompts.

SFT · Instruction
subject direction-23

Portrait Direct LoRA

Fine-tune Qwen/Qwen3.5-9B on 87 direction pairs so it coaches portrait poses from client prompts.

SFT · Instruction
dynamic range control-24

Expo Bracket LoRA

Fine-tune Qwen/Qwen3-8B on 73 bracket notes so it plans exposure brackets from contrast prompts.

SFT · Instruction
metadata tagginge-0

EXIF Schema Scribe

Fine-tune Qwen/Qwen3-8B on 80 (photo, EXIF-JSON) pairs so the model emits valid EXIF metadata from a shot description.

SFT · Structured Output
pre-production planningm-1

Shoot Plan Loom

Fine-tune Qwen/Qwen3.5-4B on 50 (brief, schedule-JSON) pairs so the model outputs a structured shoot timeline.

SFT · Structured Output
color grading rulesr-2

Grade LUT Forger

Fine-tune Qwen/Qwen3-8B on 70 (scene, LUT-JSON) pairs so the model emits a valid color grade preset.

SFT · Structured Output
photojournalism captionsr-3

Caption DSL Weaver

Fine-tune Qwen/Qwen3.5-4B on 90 (image note, caption-DSL) pairs so the model writes wire-service caption syntax.

SFT · Structured Output
composition rulesr-4

Frame Rule Carver

Fine-tune Qwen/Qwen3-8B on 60 (scene, rule-JSON) pairs so the model outputs composition guidelines as JSON.

SFT · Structured Output
gear packingt-5

Kit Pack Sheet

Fine-tune Qwen/Qwen3.5-4B on 40 (assignment, kit-JSON) pairs so the model emits a packing list schema.

SFT · Structured Output
post-processing stepsr-6

Edit Step Sequencer

Fine-tune Qwen/Qwen3-8B on 75 (raw note, edit-JSON) pairs so the model outputs an edit step sequence.

SFT · Structured Output
storyboard beatsh-7

Storyboard Beat Smith

Fine-tune Qwen/Qwen3.5-9B on 65 (concept, board-JSON) pairs so the model outputs shot-by-shot boards.

SFT · Structured Output
lighting setupr-8

Light Map Coder

Fine-tune Qwen/Qwen3-8B on 55 (mood, light-JSON) pairs so the model emits a lighting diagram schema.

SFT · Structured Output
legal releasesr-9

Model Release Writer

Fine-tune Qwen/Qwen3.5-4B on 45 (session, release-JSON) pairs so the model outputs valid release fields.

SFT · Structured Output
print preparation-10

Print Spec Drafter

Fine-tune Qwen/Qwen3-8B on 50 (image, print-JSON) pairs so the model emits print spec schemas.

SFT · Structured Output
stock metadata-11

Stock Keywordex

Fine-tune Qwen/Qwen3.5-4B on 85 (photo, keyword-JSON) pairs so the model outputs stock keyword arrays.

SFT · Structured Output
cropping rules-12

Crop Ratio Engine

Fine-tune Qwen/Qwen3-8B on 60 (frame, crop-JSON) pairs so the model emits crop ratio suggestions.

SFT · Structured Output
photo sequencing-13

Series Order Knot

Fine-tune Qwen/Qwen3.5-9B on 70 (series, order-JSON) pairs so the model outputs narrative sequence schemas.

SFT · Structured Output
analog processing-14

Recipe Card Burner

Fine-tune Qwen/Qwen3-8B on 40 (film, recipe-JSON) pairs so the model emits darkroom recipe schemas.

SFT · Structured Output
aerial planning-15

Drone Flight JSON

Fine-tune Qwen/Qwen3.5-4B on 55 (site, flight-JSON) pairs so the model outputs drone shot plans.

SFT · Structured Output
retouch regions-16

Retouch Mask Mapper

Fine-tune Qwen/Qwen3-8B on 65 (face, mask-JSON) pairs so the model emits retouch region schemas.

SFT · Structured Output
shoot permits-17

Permit Field Coder

Fine-tune Qwen/Qwen3.5-4B on 35 (location, permit-JSON) pairs so the model outputs permit field schemas.

SFT · Structured Output
optics selection-18

Lens Choice Table

Fine-tune Qwen/Qwen3-8B on 50 (scene, lens-JSON) pairs so the model emits lens recommendation schemas.

SFT · Structured Output
catalog taxonomy-19

Archive Taxonomy Root

Fine-tune Qwen/Qwen3.5-9B on 80 (collection, taxo-JSON) pairs so the model outputs archive taxonomies.

SFT · Structured Output
exposure settings-20

Exposure Triangle JSON

Fine-tune Qwen/Qwen3.5-4B on 60 (condition, exp-JSON) pairs so the model emits exposure setting schemas.

SFT · Structured Output
pose direction-21

Subject Pose Graph

Fine-tune Qwen/Qwen3-8B on 45 (brief, pose-JSON) pairs so the model outputs pose instruction schemas.

SFT · Structured Output
workflow backup-22

Backup Plan Tree

Fine-tune Qwen/Qwen3.5-4B on 40 (rig, backup-JSON) pairs so the model emits backup plan schemas.

SFT · Structured Output
image critique-23

Critique Score Card

Fine-tune Qwen/Qwen3.5-9B on 75 (photo, score-JSON) pairs so the model outputs critique score schemas.

SFT · Structured Output
project titling-24

Series Title Forge

Fine-tune Qwen/Qwen3-8B on 50 (theme, title-JSON) pairs so the model emits titled series schemas.

SFT · Structured Output
exposure parameter suggestionr-0

Aperture Whisperer

Fine-tune Qwen/Qwen3-8B on 80 scene prompts so it outputs aperture values maximizing a programmatic sharpness-depth reward.

RL · Programmatic Reward
motion blur controle-1

Shutter Cadence

Fine-tune Qwen/Qwen3.5-4B on 60 action prompts so it predicts shutter speeds scoring low unwanted-blur via grader.

RL · Programmatic Reward
noise-aware ISO choicee-2

ISO Forge

Fine-tune Qwen/Qwen3-8B on 70 low-light prompts so it selects ISO minimizing a coded noise-penalty reward.

RL · Programmatic Reward
rule-of-thirds croppingr-3

Frame Sculptor

Fine-tune Qwen/Qwen3.5-4B on 90 compositions so it emits crop boxes maximizing a geometry reward function.

RL · Programmatic Reward
color temperature tuninge-4

White Balance Sage

Fine-tune Qwen/Qwen3-8B on 75 lighting prompts so it outputs kelvin values optimizing a delta-E grader.

RL · Programmatic Reward
barrel distortion correctionx-5

Lens Distort Fix

Fine-tune Qwen/Qwen3.5-4B on 65 lens grids so it predicts correction params maximizing line-straightness reward.

RL · Programmatic Reward
focus bracketing planst-6

Focus Stack Pilot

Fine-tune Qwen/Qwen3-8B on 55 macro prompts so it outputs focus steps maximizing depth-coverage grader.

RL · Programmatic Reward
tone curve suggestionr-7

Histogram Harmonizer

Fine-tune Qwen/Qwen3.5-4B on 85 histograms so it proposes curves maximizing contrast-reward without clipping.

RL · Programmatic Reward
depth-map blur maskst-8

Bokeh Architect

Fine-tune Qwen/Qwen3-8B on 70 portraits so it outputs blur masks maximizing subject-separation grader.

RL · Programmatic Reward
exposure fusion weightsc-9

HDR Merge Tactic

Fine-tune Qwen/Qwen3.5-4B on 60 bracket sets so it weights exposures maximizing ghost-free HDR reward.

RL · Programmatic Reward
overlap estimation-10

Panorama Stitcher

Fine-tune Qwen/Qwen3-8B on 50 scene rows so it predicts overlaps maximizing seam-reward from grader.

RL · Programmatic Reward
keyword tagging rules-11

Catalog Taxonomist

Fine-tune Qwen/Qwen3.5-4B on 90 photo captions so it emits tags maximizing recall reward vs ground truth.

RL · Programmatic Reward
alt-text generation-12

Caption Curator

Fine-tune Qwen/Qwen3-8B on 80 images so it writes alt-text maximizing accessibility-reward from checker.

RL · Programmatic Reward
metadata normalization-13

EXIF Poet

Fine-tune Qwen/Qwen3.5-4B on 70 EXIF dumps so it standardizes fields maximizing schema-reward.

RL · Programmatic Reward
CMYK conversion-14

Print Gamut Guard

Fine-tune Qwen/Qwen3-8B on 65 RGB swatches so it maps tones maximizing in-gamut reward.

RL · Programmatic Reward
shot sequence planning-15

Storyboard Beats

Fine-tune Qwen/Qwen3.5-9B on 55 briefs so it orders shots maximizing narrative-reward from parser.

RL · Programmatic Reward
strobe power ratios-16

Lighting Ratio Coach

Fine-tune Qwen/Qwen3.5-4B on 75 setups so it outputs ratios maximizing modeled falloff reward.

RL · Programmatic Reward
composition emptiness-17

Negative Space Reader

Fine-tune Qwen/Qwen3-8B on 85 frames so it scores crops maximizing negative-space reward.

RL · Programmatic Reward
film grain matching-18

Grain Synthesist

Fine-tune Qwen/Qwen3.5-4B on 60 film refs so it sets grain params maximizing similarity reward.

RL · Programmatic Reward
edge darkening-19

Vignette Calibrator

Fine-tune Qwen/Qwen3-8B on 50 lenses so it predicts vignette curves maximizing naturalness reward.

RL · Programmatic Reward
look preset generation-20

Color Grade Rules

Fine-tune Qwen/Qwen3.5-9B on 90 moods so it emits LUT params maximizing style-reward from analyzer.

RL · Programmatic Reward
semantic segmentation-21

Subject Masker

Fine-tune Qwen/Qwen3-8B on 70 photos so it outputs masks maximizing IoU reward vs truth.

RL · Programmatic Reward
visible mark reduction-22

Watermark Invisibler

Fine-tune Qwen/Qwen3.5-4B on 65 marked imgs so it predicts edits maximizing conceal-reward.

RL · Programmatic Reward
action shot picking-23

Burst Selector

Fine-tune Qwen/Qwen3.5-9B on 80 bursts so it ranks frames maximizing sharpness-reward.

RL · Programmatic Reward
preset code export-24

Recipe Transcriber

Fine-tune Qwen/Qwen3-8B on 75 edits so it writes preset code maximizing parse-reward.

RL · Programmatic Reward
studio caption voicer-0

House Voice Taster

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

RL · Preference Pairs
edit suggestion tonee-1

Edit Reject Sense

Fine-tune Qwen/Qwen3.5-4B on 60 pairs so edit suggestions avoid over-processing rejected by senior editors.

RL · Preference Pairs
photojournalism captionse-2

Caption Safety Posture

Fine-tune Qwen/Qwen3-8B on 80 triples so captions respect sensitivity norms rejected by news desks.

RL · Preference Pairs
color grade languager-3

Grade Tone Whisper

Fine-tune Qwen/Qwen3.5-4B on 50 pairs so grade descriptions match muted house aesthetic over rejected vivid ones.

RL · Preference Pairs
photo essay ordere-4

Series Sequencing Taste

Fine-tune Qwen/Qwen3-8B on 70 triples so essay sequences follow narrative taste rejected by editors.

RL · Preference Pairs
crop justificatione-5

Crop Rationale Voice

Fine-tune Qwen/Qwen3.5-4B on 45 pairs so crop reasons use concise studio voice rejected by mentors.

RL · Preference Pairs
exif annotation toner-6

Exif Story Filter

Fine-tune Qwen/Qwen3.5-4B on 55 triples so exif stories avoid technical jargon rejected by clients.

RL · Preference Pairs
selection commentarym-7

Contact Sheet Skim

Fine-tune Qwen/Qwen3-8B on 65 pairs so contact notes praise keepers rejected by art directors.

RL · Preference Pairs
retouch boundary voicee-8

Retouch Ethic Line

Fine-tune Qwen/Qwen3-8B on 75 triples so retouch advice respects ethic lines rejected by agencies.

RL · Preference Pairs
photo title voicem-9

Title Poetic Calm

Fine-tune Qwen/Qwen3.5-4B on 40 pairs so titles stay calm poetic rejected by gallery curators.

RL · Preference Pairs
brief-to-caption tone-10

Prompt Caption Match

Fine-tune Qwen/Qwen3.5-4B on 50 triples so captions match brief voice rejected by creative leads.

RL · Preference Pairs
archive tagging voice-11

Archive Tag Restraint

Fine-tune Qwen/Qwen3.5-4B on 60 pairs so tags stay restrained rejected by archivists.

RL · Preference Pairs
social post voice-12

Social Caption Warmth

Fine-tune Qwen/Qwen3.5-4B on 45 triples so social captions feel warm rejected by brand managers.

RL · Preference Pairs
documentary narration-13

Doc Tone Serious

Fine-tune Qwen/Qwen3-8B on 70 triples so documentary notes stay serious rejected by filmmakers.

RL · Preference Pairs
print label voice-14

Print Label Plain

Fine-tune Qwen/Qwen3.5-4B on 35 pairs so print labels stay plain rejected by lab techs.

RL · Preference Pairs
stock keyword tone-15

Stock Keywording Calm

Fine-tune Qwen/Qwen3.5-4B on 55 triples so stock keywords avoid hype rejected by contributors.

RL · Preference Pairs
wedding album text-16

Wedding Voice Sweet

Fine-tune Qwen/Qwen3.5-4B on 50 pairs so album text stays sweet rejected by planners.

RL · Preference Pairs
street photo notes-17

Street Ethic Quiet

Fine-tune Qwen/Qwen3-8B on 65 triples so street notes stay quiet rejected by zine editors.

RL · Preference Pairs
product photo copy-18

Product Copy Crisp

Fine-tune Qwen/Qwen3.5-4B on 45 triples so product copy stays crisp rejected by clients.

RL · Preference Pairs
nature caption voice-19

Nature Calm Desc

Fine-tune Qwen/Qwen3.5-4B on 40 pairs so nature captions stay calm rejected by magazines.

RL · Preference Pairs
fashion editorial tone-20

Fashion Edge Voice

Fine-tune Qwen/Qwen3-8B on 60 triples so fashion notes keep edge rejected by editors.

RL · Preference Pairs
real estate copy-21

Real Estate Plain

Fine-tune Qwen/Qwen3.5-4B on 50 pairs so listing copy stays plain rejected by agents.

RL · Preference Pairs
food photo text-22

Food Styling Warm

Fine-tune Qwen/Qwen3.5-4B on 45 triples so food text stays warm rejected by stylists.

RL · Preference Pairs
sports caption voice-23

Sports Hype Line

Fine-tune Qwen/Qwen3-8B on 55 triples so sports captions keep hype rejected by desk.

RL · Preference Pairs
museum label tone-24

Museum Label Restraint

Fine-tune Qwen/Qwen3.5-4B on 40 pairs so labels stay restrained rejected by curators.

RL · Preference Pairs