config.toml

examples = [
    "Write a poem about AI.",
    "Write a Hello World in Rust.",
    "Explain algorithmic entropy.",
    "Write some example Markdown.",
]

default_model = "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO"

[[model]]
model_id = "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO"
name = "Nous Hermes 2 - Mixtral 8x7B - DPO"
chat_template = "ChatML"
stream = true

[[model]]
model_id = "mistralai/Mistral-7B-Instruct-v0.2"
name = "Mistral-7B-Instruct-v0.2"
chat_template = "Instruct"
stream = true

[[model]]
model_id = "mistralai/Mistral-7B-v0.1"
name = "Mistral-7B-v0.1"
stream = true

[[model]]
model_id = "mistralai/Mixtral-8x7B-Instruct-v0.1"
name = "Mixtral-8x7B"
chat_template = "Instruct"
stream = true

#[[model]]
#model_id = "codellama/CodeLlama-70b-Instruct-hf"
#chat_template = "CodeLlamaInstruct"
#stream = false

[[model]]
model_id = "TinyLlama/TinyLlama-1.1B-Chat-v1.0"
name = "TinyLlama-1.1B"
chat_template = "ChatML"
stream = false

[[model]]
model_id = "HuggingFaceH4/zephyr-7b-beta"
name = "Zephyr 7B β"
chat_template = "ChatML"
stream = true

[[model]]
model_id = "openai-community/gpt2"
name = "GPT-2"
stream = false

# note: does work (with streaming) but returns prompt
#[[model]]
#model_id = "bigcode/starcoder"
#name = "StarCoder"
#stream = false

[[model]]
model_id = "codellama/CodeLlama-7b-hf"
name = "Code Llama"
stream = true

#[[model]]
#model_id = "cognitivecomputations/TinyDolphin-2.8.1-1.1b"
#name = "TinyDolphin-2.8.1-1.1b"
#stream = false