Mettre en file une étape d'optimisation

POST

https://api.fastino.ai

training

jobs

{job_id}

optim-step

curl -X POST "https://api.fastino.ai/v1/training/jobs/$TRAINING_JOB_ID/optim-step" \
  -H "X-API-Key: $FASTINO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"adam_params": {"learning_rate": 0.0001}}'

{
  "future_id": "FUTURE_ID",
  "job_id": "SESSION_ID:train:0",
  "type": "optim_step"
}

POST /v1/training/jobs/{job_id}/optim-step Applies the gradients accumulated by forward-backward passes since the last step, using AdamW. Calls run in the order the job receives them, so you can queue this right after a forward-backward pass without waiting for it.

Request

string

requis

Your Fastino API key.

string

requis

Interactive job ID.

The JSON body is optional.

object

AdamW parameters. Omitted fields use their defaults.

Afficher properties

number

défaut:"0.0001"

Learning rate, >= 0.

number

défaut:"0.9"

From 0 up to, but not including, 1.

number

défaut:"0.95"

From 0 up to, but not including, 1.

number

défaut:"1e-12"

> 0.

number

défaut:"0"

>= 0.

number

défaut:"0"

Gradient-norm clip, >= 0. 0 disables clipping.

curl -X POST "https://api.fastino.ai/v1/training/jobs/$TRAINING_JOB_ID/optim-step" \
  -H "X-API-Key: $FASTINO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"adam_params": {"learning_rate": 0.0001}}'

Response

{
  "future_id": "FUTURE_ID",
  "job_id": "SESSION_ID:train:0",
  "type": "optim_step"
}

The completed future’s result holds the step’s metrics.

Status	Meaning
`400`	An Adam parameter is out of range, or the job was stopped.
`404`	The job doesn’t exist or belongs to another team.
`422`	The body has an unknown field.

⌘I

curl -X POST "https://api.fastino.ai/v1/training/jobs/$TRAINING_JOB_ID/optim-step" \
  -H "X-API-Key: $FASTINO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"adam_params": {"learning_rate": 0.0001}}'

{
  "future_id": "FUTURE_ID",
  "job_id": "SESSION_ID:train:0",
  "type": "optim_step"
}

Inférence

Entraînement

Mettre en file une étape d'optimisation

Request

Response

​Request

​Response

Request

Response