跳至主要内容
版本：1.1.2

Model Management

訊息

The Model Management page allows users to view and manage a list of supported models for fine-tuning or inference purposes. Models can be downloaded directly from Hugging Face repositories. Before using the feature, ensure that your device is connected to the internet, as model downloads require access to external servers.

Additionally, on the first download attempt, users will be prompted to provide an Access Token from Hugging Face. This token can be obtained by following Hugging Face's official guidelines for token generation.

Model Management

Features​

  1. Model List:
    • Displays all available models, including their names, sizes, and current processing status.
    • Users can search for specific models using the search bar.
  2. Model Details:
    • Model Name: Specifies the name of the model (e.g., meta-llama/Llama-3.1-8B-Instruct).
    • Support: Use the Phison tag for full parameters and Unsloth for the LoRA mechanism. A model tagged with both supports two different LLM fine-tuning approaches.
    • Size: Shows the storage size of each model (e.g., 29.93 GiB, 262.86 GiB).
    • Process Status: Indicates whether the model is ready to use (✓ for ready).
    • Action: Allows users to manage the model (e.g., download, delete).
  3. Download Process:
    • Displays the download progress for models that are currently being fetched.
    • Provides a clear indication of the model size and estimated time remaining.
    • Supports pause and resume functionality for interrupted downloads.
  4. Access Token Requirement:
    • During the initial download, users are required to provide a Hugging Face Access Token for authentication. Access Token
    • For token generation, visit the Hugging Face website and follow the official instructions to create and retrieve your personal token.

Supported Models​

ModelNameSupportSize
google/gemma-2b-itunsloth14.03 GiB
meta-llama/Llama-3.1-8B-Instructphison, unsloth29.93 GiB
01-ai/Yi-1.5-34B-Chatphison64.06 GiB
01-ai/Yi-1.5-9B-Chatphison16.45 GiB
HuggingFaceH4/mistral-7b-sft-betaunsloth26.98 GiB
HuggingFaceTB/SmolLM-1.7B-Instructunsloth21.10 GiB
HuggingFaceTB/SmolLM-135M-Instructunsloth1.83 GiB
HuggingFaceTB/SmolLM-360M-Instructunsloth4.67 GiB
MediaTek-Research/Breeze-7B-Instruct-v0_1phison27.90 GiB
Nexusflow/Starling-LM-7B-betaunsloth13.49 GiB
NousResearch/Hermes-2-Pro-Mistral-7Bunsloth13.49 GiB
NousResearch/Hermes-3-Llama-3.1-8Bunsloth14.97 GiB
Qwen/Qwen1.5-0.5B-Chatphison1.16 GiB
Qwen/Qwen1.5-1.8B-Chatphison3.43 GiB
Qwen/Qwen1.5-110B-Chatphison207.16 GiB
Qwen/Qwen1.5-14B-Chatphison26.40 GiB
Qwen/Qwen1.5-4B-Chatphison7.37 GiB
Qwen/Qwen1.5-72B-Chatphison134.66 GiB
Qwen/Qwen1.5-7B-Chatphison14.39 GiB
Qwen/Qwen2-0.5Bunsloth953.29 MiB
Qwen/Qwen2-0.5B-Instructunsloth953.29 MiB
Qwen/Qwen2-1.5Bunsloth2.89 GiB
Qwen/Qwen2-1.5B-Instructunsloth2.89 GiB
Qwen/Qwen2-72Bphison135.44 GiB
Qwen/Qwen2-72B-Instructphison135.44 GiB
Qwen/Qwen2-7Bphison, unsloth14.20 GiB
Qwen/Qwen2-7B-Instructunsloth14.20 GiB
Qwen/Qwen2.5-0.5B-Instructunsloth953.30 MiB
Qwen/Qwen2.5-1.5B-Instructunsloth2.89 GiB
Qwen/Qwen2.5-3B-Instructunsloth5.76 GiB
Qwen/Qwen2.5-72B-Instructphison135.44 GiB
Qwen/Qwen2.5-7B-Instructunsloth14.20 GiB
TinyLlama/TinyLlama-1.1B-Chat-v1.0unsloth2.05 GiB
akjindal53244/Llama-3.1-Storm-8Bunsloth14.97 GiB
allenai/Llama-3.1-Tulu-3-8Bunsloth14.97 GiB
deepseek-ai/DeepSeek-Coder-V2-Instructphison439.11 GiB
deepseek-ai/DeepSeek-R1-Distill-Llama-8Bunsloth14.97 GiB
deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5Bunsloth3.32 GiB
deepseek-ai/DeepSeek-R1-Distill-Qwen-7Bunsloth14.19 GiB
deepseek-ai/deepseek-llm-7b-chatphison12.88 GiB
google/codegemma-7b-itunsloth15.92 GiB
google/gemma-1.1-2b-itunsloth4.69 GiB
google/gemma-1.1-7b-itunsloth15.92 GiB
google/gemma-2-27b-itphison50.74 GiB
google/gemma-2-9b-itphison17.24 GiB
google/gemma-7b-itunsloth47.74 GiB
meta-llama/Llama-2-7b-chat-hfunsloth25.10 GiB
meta-llama/Llama-3.1-405B-Instructphison2.22 TiB
meta-llama/Llama-3.1-70B-Instructphison262.86 GiB
meta-llama/Llama-3.2-3B-Instructunsloth11.98 GiB
meta-llama/LlamaGuard-7bphison25.11 GiB
meta-llama/Meta-Llama-3-70B-Instructphison262.86 GiB
meta-llama/Meta-Llama-3-8B-Instructphison, unsloth29.93 GiB
mistralai/Mistral-7B-Instruct-v0.1unsloth27.47 GiB
mistralai/Mistral-7B-Instruct-v0.2unsloth27.47 GiB
mistralai/Mistral-7B-Instruct-v0.3unsloth27.00 GiB
mistralai/Mixtral-8x22B-Instruct-v0.1phison261.95 GiB
mistralai/Mixtral-8x7B-Instruct-v0.1phison177.40 GiB
teknium/OpenHermes-2.5-Mistral-7Bunsloth26.98 GiB
yentinglin/Llama-3-Taiwan-70B-Instructphison31.43 GiB

Steps for Access Token​

  1. Navigate to the Hugging Face website: https://huggingface.co.
  2. Log in to your Hugging Face account.
  3. Go to your account settings and select the Access Tokens tab.
  4. Generate a new token with the required permissions.
  5. Copy the token and provide it when prompted during the model download process in the application.

For more details, refer to the Hugging Face documentation.

注意

Some models, such as LLAMA, require user consent for sharing contact information before download.