How To Create Session

Step 1: Enter Session Name and Description

You must enter a unique name session (e.g., int.session_20250915155915) that satisfies the following conditions:

• Maximum 100 characters

Optional note session description with maximum 200 characters.

Step 2: Select Model

You select model from the Model Hub in two ways:

• Model Catalog: A repository of model sources from various providers such as DeepSeek, Gemma, Llama, and Qwen.
• Private Model: A repository for user-owned models and fine-tuned models. These models must already have all necessary files uploaded. Must include specific tags to indicate compatibility and readiness.

Notice: To be used in Interactive Sessions, models must meet the following criteria:

• Model type:
  • LLM: Accepts only text input
  • VLM: Accepts both text and image input
• Model size > 0
• Learning stage: Must be Instruction-tuned

Example:

Model name	Model version	Model family	Model type	Model size	Learning stage
ft_Llama-3.1-8B_20250508124054_samples-15d5e2f6fe7	15d5e2f6fe7	Llama	LLM

(→ available) | 8B

(→ available) | Instruction-tuned

(→ available) |

Step 3: Select Session Stop

You can enter and adjust the icon up-down to select appreciated time for the session. We recommend you define how the session should automatically terminate based on one or more of the following condition:

• Maximum number of interactions
• Time duration
• Model-specific limits

This ensures efficient resource usage and prevents unintended long-running sessions.

Step 4: Submit

You review and submit a session Notice: For the PAYG user To run a pipeline successfully, ensure:

• Balance has sufficient funds. Reference page: https://ai.fptcloud.com/pricing
• Create up to 20 pipelines in total
• Only 2 pipelines can be started concurrently at any given time
• Each pipeline can be configured with a maximum of 2 nodes

Notice: For the user to sign a contract To run a pipeline successfully, ensure:

• Balance has sufficient funds. Reference page: https://ai.fptcloud.com/pricing