Get model limits

Get model limits for a specific fine-tuning model.

GET /fine-tunes/models/limits

cURL

cURL

curl "https://api.together.ai/v1/fine-tunes/models/limits?model_name=meta-llama/Meta-Llama-3.1-8B-Instruct-Reference" \
     -H "Authorization: Bearer $TOGETHER_API_KEY"

import requests

url = "https://api.together.ai/v1/fine-tunes/models/limits"

headers = {"Authorization": "Bearer <token>"}

response = requests.get(url, headers=headers)

print(response.text)

const options = {method: 'GET', headers: {Authorization: 'Bearer <token>'}};

fetch('https://api.together.ai/v1/fine-tunes/models/limits', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.together.ai/v1/fine-tunes/models/limits",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "GET",
  CURLOPT_HTTPHEADER => [
    "Authorization: Bearer <token>"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

package main

import (
	"fmt"
	"net/http"
	"io"
)

func main() {

	url := "https://api.together.ai/v1/fine-tunes/models/limits"

	req, _ := http.NewRequest("GET", url, nil)

	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.get("https://api.together.ai/v1/fine-tunes/models/limits")
  .header("Authorization", "Bearer <token>")
  .asString();

require 'uri'
require 'net/http'

url = URI("https://api.together.ai/v1/fine-tunes/models/limits")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Get.new(url)
request["Authorization"] = 'Bearer <token>'

response = http.request(request)
puts response.read_body

200

{
  "model_name": "<string>",
  "lora_training": {
    "max_batch_size": 123,
    "max_batch_size_dpo": 123,
    "min_batch_size": 123,
    "max_rank": 123,
    "target_modules": [
      "<string>"
    ]
  },
  "max_num_epochs": 123,
  "max_num_evals": 123,
  "max_learning_rate": 123,
  "min_learning_rate": 123,
  "supports_full_training": true,
  "supports_vision": true,
  "supports_tools": true,
  "supports_reasoning": true,
  "merge_output_lora": true,
  "default_gradient_accumulation_steps": 123,
  "max_num_checkpoints": 123,
  "min_max_seq_length": 123,
  "max_seq_length_sft": 123,
  "max_seq_length_dpo": 123,
  "full_training": {
    "max_batch_size": 123,
    "max_batch_size_dpo": 123,
    "min_batch_size": 123
  }
}

{
  "message": "<string>"
}

Authorizations

Authorization · string · header · default: default · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Query Parameters

model_name · string · required

The model name to get limits for.

Response

200

application/json

Model limits.

Model limits for fine-tuning.

model_name · string · required

The name of the model.

lora_training · object · required

Limits for LoRA training.

Show child attributes

max_num_epochs · integer · required

Maximum number of training epochs.

max_num_evals · integer · required

Maximum number of evaluations.

max_learning_rate · number · required

Maximum learning rate.

min_learning_rate · number · required

Minimum learning rate.

supports_full_training · boolean · required

Whether the model supports full (non-LoRA) fine-tuning. When false, only LoRA fine-tuning is available and the full_training limits are reported as zero.

supports_vision · boolean · required

Whether the model supports vision/multimodal inputs.

supports_tools · boolean · required

Whether the model supports tool/function calling.

supports_reasoning · boolean · required

Whether the model supports reasoning.

merge_output_lora · boolean · required

Whether a merged checkpoint (the base model with the trained LoRA adapter fused in) is produced for LoRA fine-tunes of this model, in addition to the standalone adapter.

default_gradient_accumulation_steps · integer · required

Default gradient accumulation steps used when a fine-tune request omits the value or sets it to 0.

max_num_checkpoints · integer · required

Maximum number of checkpoints that can be saved during a fine-tuning job.

min_max_seq_length · integer · required

Minimum value allowed for the max_seq_length hyperparameter.

max_seq_length_sft · integer · required

Maximum sequence length supported for SFT training.

max_seq_length_dpo · integer · required

Maximum sequence length supported for DPO training.

full_training · object

Limits for full training.

Show child attributes