Alibaba Cloud Model Studio:My Models

Manage models fine-tuned in Alibaba Cloud Model Studio and models imported after local fine-tuning (LoRA fine-tuning or full-parameter fine-tuning only), covering first-time OSS authorization, model file preparation and validation, import form completion, model status management, and common issue troubleshooting. These models support cloud deployment and provide high-concurrency, low-latency inference services.

Supported Models

On the My Models page, you can manage the following two types of models:

• Fine-tuned models: Models fine-tuned in Alibaba Cloud Model Studio. For more information, see Supported models for model fine-tuning.

• Imported models: Models imported from Object Storage Service (OSS). This supports importing models that underwent LoRA fine-tuning or full-parameter fine-tuning.

  Importing full-parameter fine-tuned models is a whitelist feature. To enable it, contact your account manager.

  Model list:

  China (Beijing)

  Supported Models	Model Name
  Qwen3	Qwen3-32B
  Qwen3	Qwen3-14B
  Qwen3	Qwen3-8B
  Qwen3	Qwen3-4B-Instruct-2507
  Qwen3-VL	Qwen3-VL-8B-Instruct
  Qwen2.5	Qwen2.5-72B-Instruct
  Qwen2.5	Qwen2.5-32B-Instruct
  Qwen2.5	Qwen2.5-14B-Instruct
  Qwen2.5	Qwen2.5-7B-Instruct
  Qwen2.5-VL	Qwen2.5-VL-72B-Instruct
  Qwen2.5-VL	Qwen2.5-VL-7B-Instruct

  Singapore

  Supported Models	Model Name
  -	-

  Data updates may be delayed. Refer to the actual display on the interface when you import models.

Import Models

This section describes how to import models from OSS using the Alibaba Cloud Model Studio console.

Importing models is not supported through API or command line operations.

1. On the My Models page, click Import Model to open the Import Model interface.

2. Enter a Model Name and select the Base Model (the model to be fine-tuned).

3. For Import Method, select Import from OSS. Currently, no other methods are supported. You need to manually upload the associated files for your LoRA fine-tuned or full-parameter fine-tuned model to Alibaba Cloud Object Storage Service (OSS). Upload View model file examples

   Important

   • For the first import, complete the authorization as prompted on the interface and add a tag to the target bucket. For more information, see How do I import files from OSS to Model Studio for the first time?.
   • Supported OSS bucket storage classes do not include Archive, Cold Archive, or Deep Cold Archive. Encrypted buckets are supported. Private buckets are supported.
   • Accessing files in the root directory of an OSS bucket is not supported. Select an existing subdirectory or create a new one within the OSS bucket for Alibaba Cloud Model Studio to access.
   • You can import model files of any size. After import, they use the free storage space provided by Alibaba Cloud Model Studio.

4. After clicking OK, the system begins importing the model. During peak request times, this process may take longer. You can wait patiently.

Prepare LoRA Model Files

Before importing, place the LoRA model files in a subdirectory of an OSS bucket (the bucket root directory is not supported) according to the requirements below, and pass the system's automatic validation before submission. The model files must be placed directly under the selected subdirectory; the system detects them automatically.

Required files and directory structure

The subdirectory must contain the following files: adapter_model.safetensors (the LoRA adapter weights in SafeTensors format), adapter_config.json (the configuration file containing parameters such as rank and alpha), and config.json (the base model configuration). After you select a directory, the system automatically validates the format and integrity of these files.

Training parameter constraints

• Rank values: The rank must be one of 8, 16, 32, or 64, and all LoRA layers of the same model must use the same rank value; otherwise the model cannot be imported.
• Vocabulary must not be modified: Models that added new tokens or modified the original vocabulary during training cannot be imported; the vocabulary must exactly match that of the base model.
• Chat template must not be modified: Models that modified the chat_template during training cannot be imported; the chat_template must match the base model's default configuration. The chat_template is located in the chat_template field of config.json or tokenizer_config.json.
• Vision-language models must freeze VIT: Vision-language models must freeze the Vision Transformer part. If the LoRA adapter contains visual-related weight parameters (i.e., VIT is not frozen), the model cannot be imported.

You can run the following script before import to check whether adapter_model.safetensors contains parameter keys starting with visual, to determine whether VIT is frozen.

from safetensors import safe_open
import argparse

def print_safetensor_structure(file_path):
    print(f"Loading safetensor file: {file_path}")
    print("="*80)

    with safe_open(file_path, framework="pt") as f:
        keys = f.keys()
        print(f"Found {len(keys)} tensors in the file:\n")

        for key in sorted(keys):
            tensor = f.get_tensor(key)
            shape = tuple(tensor.shape)
            dtype = str(tensor.dtype)
            device = tensor.device if hasattr(tensor, 'device') else 'cpu'

            lora_tag = " [LoRA]" if "lora_A" in key or "lora_B" in key else ""

            print(f"[{dtype:>14}] {shape} | {key} {lora_tag}")

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Print structure of a .safetensors LoRA adapter.")
    parser.add_argument("filepath", type=str, help="Path to the .safetensors file")
    args = parser.parse_args()

    print_safetensor_structure(args.filepath)

Method: If the script output contains parameter keys starting with visual (for example, visual.encoder.layer.0...), the VIT part is not frozen and the model cannot be imported; if only LoRA-related parameter keys such as lora_A and lora_B are present, VIT is frozen and the model can be imported normally.

Pre-submission automatic validation

After selecting a model directory on the Import Model page, the system automatically calls the file validation API to check the format and integrity of the model files in the directory. If validation fails, a red prompt appears under the directory field and blocks submission; correct the files per the prompt before submitting. A common failure cause is missing required files, corresponding to error code AvailableModelFileNotFound.

Import Form Fields

The import model form fields are as follows:

Imported model states include Creating (importing), Created (ready to deploy), Creation Failed (import failed), and Expired (source files changed).

What to do next

Manage My Models

On the My Models page, you can view all fine-tuned and imported models in the current workspace. You can also perform operations such as deployment and deletion.

Delete: Undeploy models before deleting them. If the model comes from Model Fine-tuning, this operation does not delete records from the Model Fine-tuning interface.

Status and transitions

Model states include Creating, Created, Creation Failed, and Expired. Creating means the import is in progress; Created means the import is complete and the model is ready for deployment; Creation Failed means the import did not succeed; Expired means the OSS source model files changed after the model was successfully created. The list silently auto-refreshes every 3 seconds for models in the Creating state; this is normal behavior, not an API exception.

Next to the Creation Failed status is a "Details" link; hover over it to view the failure error code (such as AvailableModelFileNotFound) and the corresponding oss://bucket/path path, to help locate the failed file.

For the Expired status, hover to view a pop-up showing "The following files have detected updates" and the list of changed source file names, prompting you to re-import.

Actions column availability

Each row's actions column provides three buttons: Deploy, Incremental Training, and Delete. Button availability depends on the model status:

• Deploy: Available only in the Created state; clicking it opens the deployment creation page. For deployment operations, see Deployment and O&M. In other states, or when no deployment method is available, the button is disabled.
• Incremental Training: Models imported from OSS do not support incremental training; the button is disabled.
• Delete: Models in the Creating state cannot be deleted; quantized models must be deleted from the model compression page.

After clicking Delete in the actions column and confirming, only the model record on the Model Studio side is removed; OSS source files are not affected. The search box at the top of the list filters models by name and supports clearing/resetting. The Source column renders by import source: OSS imports show oss://bucket/path, training jobs show the source job ID (or "Training job deleted" if deleted). The Supported Deployment Methods column displays the deployment methods and training method tags (such as full-parameter, LoRA, or quantization tags), as well as billing methods (such as per-token billing or per-model-unit billing); when no deployment method is available, it shows "-".

All operations are performed per single model; batch deletion or batch deployment is not supported.

Invoke My Models

Models provide inference services only after successful deployment (API invocation only). For specific operations and billing methods, see Invoke After Deployment.

• Key restrictions: After you successfully deploy models fine-tuned in Alibaba Cloud Model Studio, invoke them only using the API key of their workspace. Currently, only DashScope invocation is supported, not (OpenAI compatible) invocation.
• Optimize inference results: When invoking, explicitly configure inference hyperparameters (such as presence_penalty, frequency_penalty, and repetition_penalty). Ensure these settings match the inference hyperparameters used during your offline fine-tuning.

Model File Examples

• LoRA fine-tuning typically generates two checkpoint files: best_model_checkpoint (the checkpoint with the best performance on the validation set) and last_model_checkpoint (the last saved checkpoint). You can import either one as needed.

  Ensure the imported model files include adapter_model.safetensors and adapter_config.json. To learn about LoRA fine-tuning methods, refer to the model fine-tuning section in the Alibaba Cloud Large Language Model ACP Course.

  The checkpoint file directory typically contains the following files: global_step100/ (subdirectory), README.md, adapter_config.json, adapter_model.safetensors, additional_config.json, configuration.json, generation_config.json, latest, rng_state_0.pth, and rng_state_1.pth.

• Files to import for full-parameter fine-tuning must be consistent with open source models.

FAQ

{
    "Action": [
        "ram:CreateServiceLinkedRole"
    ],
    "Resource": "*",
    "Effect": "Allow",
    "Condition": {
        "StringEquals": {
            "ram:ServiceName": "datahub.sfm.aliyuncs.com"
        }
    }
}