Kush/upgrade changes #60
Conversation
✅ Deploy Preview for alloradocs ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
pages/devs/workers/walkthroughs/walkthrough-price-prediction-worker.mdx
Outdated
Show resolved
Hide resolved
| @@ -38,6 +38,10 @@ type MsgCreateNewTopic struct { | |||
| AllowNegative bool `json:"allow_negative,omitempty"` | |||
| // the numerical precision at which the network should distinguish differences in the logarithm of the loss | |||
| Epsilon github_com_allora_network_allora_chain_math.Dec `json:"epsilon"` | |||
| MeritSortitionAlpha github_com_allora_network_allora_chain_math.Dec `json:"merit_sortition_alpha"` | |||
There was a problem hiding this comment.
There are comments in the other fields, seems proper to add them here too.
There was a problem hiding this comment.
MeritSortitionAlpha:
// alpha parameter to weight previous score EMA importance in filtering the active set of inferers, forecasters, and reputers
ActiveXQuantile
// marks the quantile of lower Xs by score on the active set that may be replaced by new Xs each epoch
| "loss-calculation-eth.wasm" \ # LossMethod | ||
| "bafybeigpiwl3o73zvvl6dxdqu7zqcub5mhg65jiky2xqb4rdhfmikswzqm" \ # InferenceLogic | ||
| "allora-inference-function.wasm" \ # InferenceMethod | ||
| "mse" \ # LossMethod | ||
| 3600 \ # EpochLength | ||
| 0 \ # GroundTruthLag |
There was a problem hiding this comment.
Ground Truth lag should be at least EpochLength. Using EpochLength lgtm as an example here. Most if not all of our topics are defined like that.
| 3600 \ # EpochLength | ||
| 0 \ # GroundTruthLag | ||
| "ETH" # DefaultArg | ||
| 3 \ # WorkerSubmissionWindow |
There was a problem hiding this comment.
worker submission window suggested to be 12 at least.
This is 12 blocks of time we leave workers to make submissions.
As this is an example this can be set to 36 safely (1% of the EpochLength).
No description provided.