Ruka hadi kwenye maudhui
Kiswahili
Pata ufunguo wa API

LTX 2.3 Trainer (V2) - Keyframe Interpolation

Unganisha

Kitambulisho cha modeli ltx23-trainer-v2/interpolate 
# Run the model; wait up to 20 s for the finished result in the same call
curl -X POST "https://api.doitong.com/v1/run/ltx23-trainer-v2/interpolate?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "training_data_url": "https://example.com/your-file.zip"
  }'

# Still "queued" or "running"? Poll with the id from the response:
curl "https://api.doitong.com/v1/requests/REQUEST_ID?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY"

Ombi hili: $ 4.80 salio za API 4,800 Iendeshe kwenye ukurasa huu

Pata ufunguo wa API
Lightricks Mafunzo 1–60 s

Hufunza LoRA inayozalisha mwendo kati ya fremu muhimu ya kwanza na ya mwisho, kwa hiari ikiwa na fremu ya kati

OpenAPI Markdown
Lightricks · Faili

Tokeo lako la kwanza linaonekana hapa

Unda ombi hapa chini na uliendeshe kwa ufunguo wako wa API. Matokeo yanakiliwa kwenye hifadhi yetu, yakihifadhiwa milele au 7 siku bila malipo.

Unda ombi

Vidhibiti vinazalishwa kutoka kwenye schema ya modeli. Msimbo na bei husasishwa unapovibadilisha.

resolution chaguomsingi medium

Mwonekano wa mafunzo. Kuongeza mwonekano huongeza mahitaji ya kumbukumbu.

aspect_ratio chaguomsingi 1:1

Uwiano wa mafunzo.

•inahitajika

URL ya `.zip` iliyo na data ya mafunzo, iliyopangwa kulingana na hali iliyochaguliwa ya mafunzo. Kila faili ya media inaweza kuwa na manukuu ya `.txt` yenye jina msingi linalolingana.

chaguomsingi 2000

Jumla ya idadi ya hatua za mafunzo.

Ya kina validation_negative_prompt · rank · validation_resolution · validation_aspect_ratio · trigger_phrase · learning_rate · number_of_frames · frame_rate · split_input_duration_threshold · validation_number_of_frames · validation_frame_rate · stg_scale · auto_scale_input · split_input_into_scenes · debug_dataset · include_middle_keyframe · validation
chaguomsingi blurry, out of focus,...

Kidokezo hasi kinachotumiwa wakati wa uthibitishaji, ambao hutumia mchakato wa hakikisho wa hatua moja. Kwa kuwa mchakato huo hukadiria uelekezaji wa uzalishaji uliosafishwa wa hatua mbili, uelekezaji na ubora wa hakikisho hutofautiana na matokeo ya mwisho.

rank chaguomsingi 32

Daraja la urekebishaji la LoRA, linalobadilisha matumizi makubwa ya kumbukumbu kwa uwezo mkubwa zaidi kadiri mpangilio unavyoongezeka.

validation_resolution chaguomsingi high

Mwonekano unaotumiwa wakati wa uthibitishaji.

validation_aspect_ratio chaguomsingi 1:1

Uwiano unaotumiwa wakati wa uthibitishaji.

chaguomsingi

Kauli ya uanzishaji wa mtindo wa LoRA, inayowekwa mwanzoni mwa manukuu wakati wa mafunzo.

chaguomsingi 0.0002

Kasi ya ujifunzaji ya uboreshaji.

chaguomsingi 89

Idadi ya fremu kwa kila sampuli ya mafunzo. Kugawa idadi hiyo kwa 8 lazima kuache salio la 1; mifano ni 1, 9, 17, 25, 33, 41, 49, 57, 65, 73, 81, 89, 97.

chaguomsingi 24

Kasi lengwa ya fremu ya video ya mafunzo kwa fremu kwa sekunde. Kasi asilia ya LTX-2.3 ni 24.

chaguomsingi 30

Kichocheo cha kugawa mandhari kinachoonyeshwa kama muda kwa sekunde: video ndefu kuliko kizingiti pekee ndizo hugawanywa.

chaguomsingi 89

Idadi ya fremu kwa kila video ya uthibitishaji.

chaguomsingi 24

Kasi lengwa ya fremu za video za uthibitishaji, kwa fremu kwa sekunde; LTX-2.3 huendesha asili kwa 24.

chaguomsingi 1

Kipimo cha Mwongozo wa Anga na Muda (STG), ambapo 1.0 inapendekezwa na 0.0 huzima STG.

chaguomsingi []

Data ya uthibitishaji iliyotolewa kama orodha; ingizo linaunganisha picha za keyframe na kidokezo.

POST /run/ltx23-trainer-v2/interpolate?wait=20 makadirio → $ 4.80 
# Run the model; wait up to 20 s for the finished result in the same call
curl -X POST "https://api.doitong.com/v1/run/ltx23-trainer-v2/interpolate?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "training_data_url": "https://example.com/your-file.zip"
  }'

# Still "queued" or "running"? Poll with the id from the response:
curl "https://api.doitong.com/v1/requests/REQUEST_ID?wait=20" \
  -H "Authorization: Bearer $DOITONG_API_KEY"

Endesha ombi hili

Key yako inabaki kwenye kivinjari hiki: inatumwa kwa Doitong API pekee na kuhifadhiwa (localStorage) tu ukiweka tiki kwenye Kumbuka. Pata ufunguo wa API

Kuhusu LTX 2.3 Trainer (V2) - Keyframe Interpolation

Hufunza LoRA kwa interpolation ya fremu muhimu: wakati wa inferensi unatoa fremu za kwanza na za mwisho, kisha modeli huzalisha video iliyo katikati. Washa `include_middle_keyframe` ili kufunza kwa fremu ya kwanza, ya kati na ya mwisho; kila sampuli ya uthibitishaji lazima itoe `middle_image_url`. Data huja kama zip kupitia `training_data_url`, ikiwa na hadi maingizo 2 ya uthibitishaji.

Vigezo

JinaAina ChaguomsingiInaruhusiwa Maelezo
training_data_urlinahitajika
Ainastring
URL ya `.zip` iliyo na data ya mafunzo, iliyopangwa kulingana na hali iliyochaguliwa ya mafunzo. Kila faili ya media inaweza kuwa na manukuu ya `.txt` yenye jina msingi linalolingana.
ranksi lazima
Ainaenum
Chaguomsingi32
8163264128
Daraja la urekebishaji la LoRA, linalobadilisha matumizi makubwa ya kumbukumbu kwa uwezo mkubwa zaidi kadiri mpangilio unavyoongezeka.
number_of_stepssi lazima
Ainainteger
Chaguomsingi2000
100 – 10,000
Jumla ya idadi ya hatua za mafunzo.
learning_ratesi lazima
Ainanumber
Chaguomsingi0.0002
0.000001 – 1
Kasi ya ujifunzaji ya uboreshaji.
number_of_framessi lazima
Ainainteger
Chaguomsingi89
9 – 121
Idadi ya fremu kwa kila sampuli ya mafunzo. Kugawa idadi hiyo kwa 8 lazima kuache salio la 1; mifano ni 1, 9, 17, 25, 33, 41, 49, 57, 65, 73, 81, 89, 97.
frame_ratesi lazima
Ainainteger
Chaguomsingi24
8 – 60
Kasi lengwa ya fremu ya video ya mafunzo kwa fremu kwa sekunde. Kasi asilia ya LTX-2.3 ni 24.
resolutionsi lazima
Ainaenum
Chaguomsingimedium
lowmediumhigh
Mwonekano wa mafunzo. Kuongeza mwonekano huongeza mahitaji ya kumbukumbu.
aspect_ratiosi lazima
Ainaenum
Chaguomsingi1:1
16:91:19:16
Uwiano wa mafunzo.
trigger_phrasesi lazima
Ainastring
Chaguomsingi
Kauli ya uanzishaji wa mtindo wa LoRA, inayowekwa mwanzoni mwa manukuu wakati wa mafunzo.
auto_scale_inputsi lazima
Ainaboolean
Chaguomsingifalse
Marekebisho ya video kiotomatiki ili yalingane na FPS na idadi ya fremu iliyoombwa yanapowashwa; seti za data za picha hazibadilishwi.
split_input_into_scenessi lazima
Ainaboolean
Chaguomsingitrue
Ugawaji wa mandhari kwa video zinazozidi kizingiti cha muda, ukiwashwa wakati true.
split_input_duration_thresholdsi lazima
Ainanumber
Chaguomsingi30
1 – 60
Kichocheo cha kugawa mandhari kinachoonyeshwa kama muda kwa sekunde: video ndefu kuliko kizingiti pekee ndizo hugawanywa.
debug_datasetsi lazima
Ainaboolean
Chaguomsingifalse
Hali ya ukaguzi wa uchakataji awali kwa kuangalia video, picha na manukuu kabla ya uendeshaji kamili wa mafunzo. Kuiwasha humfanya mkufunzi arudishe seti ya data ya mafunzo iliyochakatwa katika jalada linaloweza kupakuliwa kwa ukaguzi wa mikono.
validationsi lazima
Ainaobject[]
Chaguomsingi[]
≤ vipengee 2
Data ya uthibitishaji iliyotolewa kama orodha; ingizo linaunganisha picha za keyframe na kidokezo.
validation[].promptinahitajika
Ainastring
Kidokezo kinachotumiwa wakati wa uthibitishaji.
validation[].start_image_urlinahitajika
Ainaurl · image
≥ 1 characters
URL ya picha ya keyframe kwa fremu ya mwanzo.
validation[].end_image_urlinahitajika
Ainaurl · image
≥ 1 characters
URL ya picha ya keyframe kwa fremu ya mwisho.
validation[].middle_image_urlsi lazima
Ainaurl · image
URL ya picha ya keyframe ya kati kwa interpolation ya first+middle+last. Uendeshaji wenye `include_middle_keyframe=true` huuhitaji; vinginevyo ni ya hiari.
validation_negative_promptsi lazima
Ainastring
Chaguomsingiblurry, out of focus, overexposed, underexposed, low contrast, washed out colors, excessive noise, grainy texture, poor lighting, flickering, motion blur, distorted proportions, unnatural skin tones, deformed facial features, asymmetrical face, missing facial features, extra limbs, disfigured hands, wrong hand count, artifacts around text, inconsistent perspective, camera shake, incorrect depth of field, background too sharp, background clutter, distracting reflections, harsh shadows, inconsistent lighting direction, color banding, cartoonish rendering, 3D CGI look, unrealistic materials, uncanny valley effect, incorrect ethnicity, wrong gender, exaggerated expressions, wrong gaze direction, mismatched lip sync, silent or muted audio, distorted voice, robotic voice, echo, background noise, off-sync audio, incorrect dialogue, added dialogue, repetitive speech, jittery movement, awkward pauses, incorrect timing, unnatural transitions, inconsistent framing, tilted camera, flat lighting, inconsistent tone, cinematic oversaturation, stylized filters, or AI artifacts.
Kidokezo hasi kinachotumiwa wakati wa uthibitishaji, ambao hutumia mchakato wa hakikisho wa hatua moja. Kwa kuwa mchakato huo hukadiria uelekezaji wa uzalishaji uliosafishwa wa hatua mbili, uelekezaji na ubora wa hakikisho hutofautiana na matokeo ya mwisho.
validation_number_of_framessi lazima
Ainainteger
Chaguomsingi89
9 – 121
Idadi ya fremu kwa kila video ya uthibitishaji.
validation_frame_ratesi lazima
Ainainteger
Chaguomsingi24
8 – 60
Kasi lengwa ya fremu za video za uthibitishaji, kwa fremu kwa sekunde; LTX-2.3 huendesha asili kwa 24.
validation_resolutionsi lazima
Ainaenum
Chaguomsingihigh
lowmediumhigh
Mwonekano unaotumiwa wakati wa uthibitishaji.
validation_aspect_ratiosi lazima
Ainaenum
Chaguomsingi1:1
16:91:19:16
Uwiano unaotumiwa wakati wa uthibitishaji.
stg_scalesi lazima
Ainanumber
Chaguomsingi1
0 – 3
Kipimo cha Mwongozo wa Anga na Muda (STG), ambapo 1.0 inapendekezwa na 0.0 huzima STG.
include_middle_keyframesi lazima
Ainaboolean
Chaguomsingifalse
Kujumuisha keyframe ya kati kwa uzalishaji wa video: true huhitaji `middle_image_url` katika kila sampuli ya uthibitishaji na huhifadhi fremu ya kati pamoja na ya kwanza na ya mwisho, badala ya kutumia ya kwanza na ya mwisho pekee.
inahitajika

Tokeo

GET /requests/req_8k2m4q7x1z9c3v5b6n0p2r4t
{
  "id": "req_8k2m4q7x1z9c3v5b6n0p2r4t",
  "object": "request",
  "model": "ltx23-trainer-v2/interpolate",
  "status": "succeeded",
  "created_at": "2026-10-20T12:00:00Z",
  "started_at": "2026-10-20T12:00:02Z",
  "completed_at": "2026-10-20T12:00:09Z",
  "queue_position": null,
  "cost": {
    "credits": 4800,
    "usd": "4.80",
    "charged": 4800,
    "refunded": 0
  },
  "input": {
    "training_data_url": "https://example.com/your-file.zip"
  },
  "output": {
    "video": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "audio": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "lora_file": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "config_file": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    },
    "debug_dataset": {
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_name": "0-output.jpeg",
      "file_size": 1843200
    }
  },
  "outputs": [
    {
      "type": "file",
      "url": "https://content.doitong.com/dev-api/2026/10/20/req_8k2m4q7x1z9c3v5b6n0p2r4t/0-output.jpeg",
      "content_type": "image/jpeg",
      "file_size": 1843200,
      "stored": true
    }
  ],
  "error": null,
  "metrics": {
    "inference_seconds": 6.8
  },
  "urls": {
    "self": "https://api.doitong.com/v1/requests/req_8k2m4q7x1z9c3v5b6n0p2r4t",
    "cancel": "https://api.doitong.com/v1/requests/req_8k2m4q7x1z9c3v5b6n0p2r4t/cancel"
  }
}
Faili zako, chaguo lakoKila outputs[].url ni nakala kwenye hifadhi yetu: huhifadhiwa milele (hutozwa kwa GB kwa mwezi) au 7 siku bila malipo, kulingana na mpangilio wa akaunti yako au ?storage=.
Muundo uleule kwenye kila modelitype, url, content_type, width, height na duration, kwa picha, video na sauti. output ni kitu cha matokeo cha modeli yenyewe.
Sehemu za tokeo
video
object {url, content_type, file_name, file_size} | nullHakikisho la video lililounganishwa kutoka kwa uthibitishaji, linapopatikana kwa endpoint zinazozalisha video.
audio
object {url, content_type, file_name, file_size} | nullSauti iliyounganishwa kutoka kwa uthibitishaji, inayotolewa kama hakikisho inapopatikana kwa njia hizi za sauti pekee: /audio-inpaint, /audio-extend-suffix, /audio-extend-prefix, /t2a, /a2a, na /v2a.
lora_file
object {url, content_type, file_name, file_size}URL ya uzito wa LoRA uliofunzwa katika umbizo la .safetensors.
config_file
object {url, content_type, file_name, file_size}Usanidi wa kuweka endpoint ya uelekezaji.
debug_dataset
object {url, content_type, file_name, file_size} | nullTokeo la utatuzi linalopatikana `debug_dataset` inapowashwa: jalada linaloweza kupakuliwa na lenye data ya mafunzo baada ya uchakataji awali.

Miundo zaidi: Mafunzo

Kwa mawakala wa AI: ukurasa huu katika Markdown unapatikana kwenye https://doitong.com/sw/developers/models/ltx23-trainer-v2/interpolate.md. Modeli na miongozo yote: https://doitong.com/developers/llms.txt