Edit model card
YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/model-cards#model-card-metadata)

πŸ’¬ Join us in Chai Prize Competition Discord

Model description

phase2_winner_13b was the winner model of Phase2 Chai Prize Competition, based on Llama-2-13b.

Submit model to Chai Prize Competition

Install the Chai Guanaco package:

pip install chai-guanaco

For one-off authentication run the following in your terminal:

chai-guanaco login

Submit model:

import chai_guanaco as chai

model_url = "ChaiML/phase2_winner_13b2"

generation_parameters = {
    'temperature': 0.8,
    'top_p': 0.2,
    "top_k": 40,
    "stopping_words": ['\n'],
    "presence_penalty": 0.,
    "frequency_penalty": 0.,
    }

submission_parameters = {
    'model_repo': model_url,
    'generation_params': generation_parameters,
    }

submitter = chai.ModelSubmitter()
submission_id = submitter.submit(submission_parameters)

Prompt formatting

from chai_guanaco.formatters import PromptFormatter

class PygmalionFormatter(PromptFormatter):
    memory_template = "{bot_name}'s Persona: {memory}\n####\n"
    prompt_template = "{prompt}\n<START>\n"
    bot_template = "{bot_name}: {message}\n"
    user_template = "{user_name}: {message}\n"
    response_template = "{bot_name}:"

Example of model input:

{bot_name}'s Persona: {memory}
####
{prompt}
<START>
[Historical conversations]
{bot_name}:
Downloads last month
14
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.