Implement BuyYes and BuyNo functions with Omen API #54
Conversation
|
Important Auto Review SkippedAuto reviews are disabled on base/target branches other than the default branch. Please add the base/target branch pattern to the list of additional branches to be reviewed in the settings. Please check the settings in the CodeRabbit UI or the To trigger a single review, invoke the Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media? TipsChatThere are 3 ways to chat with CodeRabbit:
Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (invoked as PR comments)
Additionally, you can add CodeRabbit Configration File (
|
| limit=5, | ||
| ) | ||
|
|
||
| markets: list[OmenAgentMarket] = get_omen_binary_markets() | ||
| market_questions_and_prices = [] | ||
| for market in markets: | ||
| market_questions_and_prices.append(market.question) |
There was a problem hiding this comment.
Maybe just a style preference, but I would prefer having some kind of tuple structure, like [(market_price, market_question)] instead of a list containing different item types. What do you think?
There was a problem hiding this comment.
Agreed, is there a reason why we can't use some Pydantic model here?
This seems like it could work:
from pydantic import BaseModel
class MicroMarket(BaseModel):
question: str
p_yes: float
def __str__(self) -> str:
return f"'{self.question}' with probability of yes: {self.p_yes:.2%}"
mm = MicroMarket(question="Will it rain tomorrow?", p_yes=0.7)
prompt = f"""
You have following markets:
- {mm}
"""
print(prompt)prints:
> python test.py [9:44:12]
You have following markets:
- 'Will it rain tomorrow?' with probability of yes: 70.00%There was a problem hiding this comment.
Nice, will do this!
I think this is still a starting point though, and we might want to do some thinking/experimenting how we want the api to look like for the agent. We could expose a bunch more things (opening time, closing time, liquidity, ...), either as extra fields of this pydantic class, or as separate Functions the agent can call.
My hunch is that separate functions makes more sense for token efficiency, but need to check.
Also, we could expose just the market Id here, and have the agent pass that into the other getter functions, so that we don't need get_omen_market_by_title here https://github.com/gnosis/prediction-market-agent-tooling/pull/172/files.
| return ["Will Joe Biden get reelected in 2024?", 2.3] | ||
|
|
||
| def __call__(self, market: str, amount: float) -> str: | ||
| if self.outcome == "yes": |
There was a problem hiding this comment.
Outcomes on Omen are Yes and No currently, how come it's lowercased here? We have also constants for it:
There was a problem hiding this comment.
Good point, fixed.
|
I see it currently uses Omen-specific classes/functions a lot, are you planning on having this deployable against Manifold as well? As we do with others? |
Yes, would like to support all markets we support. Once we figure out the set of Functions the agent needs, we can make them members of the I would say that there isn't the same incentive to use manifold here because it's free, as for this agent I think token cost >> betting cost |
f6d5e5c
into
48-first-steps-towards-general-agent-1n
* Added getMarkets function and passed it to agent * First commit for microchain agent (#52) * First commit with microchain agent * Separate out functions to separate file * Mypy and formatting * Formatting * Use branch of microchain that contains token-tracking functionality (#53) * Use branch of microchain that contains token-tracking functionality * Formatting * Review comment * Implement BuyYes and BuyNo functions with Omen API (#54) * Implement BuyYes and BuyNo functions with Omen API * Remove 'general_agent' --------- Co-authored-by: Gabriel Fior <fior.gbm@gmail.com>
No description provided.