LLM Farm 1.0.1

LLMFarm is an application available for iOS and MacOS designed for managing large language models (LLM). It provides the functionality to load various LLMs customizable with specific parameters.

Features:

• Multiple inference capabilities
• Diverse sampling methods
• Metal integration
• Model configuration templates
• Support for LoRA adapters and LoRA FineTune and Export features

Inferences:

• LLaMA
• GPTNeoX
• Replit
• GPT2 + Cerebras
• Starcoder (Santacoder)
• RWKV (20B tokenizer)
• Falcon
• MPT
• Bloom
• StableLM-3b-4e1t
• Qwen

Multimodal:

• LLaVA 1.5 models
• Obsidian
• MobileVLM 1.7B/3B models

Note: For detailed model settings for Falcon, Alpaca, GPT4All, Chinese LLaMA/Alpaca, Chinese LLaMA-2/Alpaca-2, Vigogne (French), Vicuna, Koala, OpenBuddy (Multilingual), Pygmalion/Metharme, WizardLM, Baichuan 1 & 2 + derivatives, Aquila 1 & 2, Mistral AI v0.1, Refact, Persimmon 8B, MPT, Bloom, please select the llama inference in the model settings.

Sampling methods:

• Temperature variations (temp, tok-k, top-p)
• Tail Free Sampling (TFS)
• Locally Typical Sampling
• Mirostat
• Greedy sampling
• Grammar-based sampling (does not work with GGJTv3)