Models
The models on the network
The judge and the market model: inputs, outputs, roots, step counts, windows and bonds.
Two models are registered on the network: judge, which answers a claim against a passage of evidence, and market, which reads the tone of a market headline or statement. This page gives what each one takes, what it answers, and every figure that identifies it.
judge
Reads a claim and a passage of evidence. Answers yes, no or unclear.
| Source weights | gokulsrinivasagan/bert_uncased_L-2_H-256_A-4_mnli, Apache-2.0 |
| Architecture | BERT, 2 layers, width 256, 4 heads, trained on MNLI |
| Parameters | 9,591,811 |
| Input | [CLS] evidence [SEP] claim [SEP] in uncased WordPiece, at most 64 tokens |
| Output | 1 token: the label |
| Steps | 134,491 |
| State pages | 1,414 |
| Weight chunks | 64,512 |
The output token is the index of the label:
| Token | Label |
|---|---|
0 | yes |
1 | unclear |
2 | no |
The SDK exports the same list as JUDGE_LABELS.
Asking the judge takes a claim and its evidence:
const { receipt } = await laplace.ask('judge', {
claim: 'The bridge opened in 1932.',
evidence: 'The bridge opened to traffic in March 1932.',
})When the two texts together pass 64 tokens, the longer one loses tokens from its end until both fit.
Roots
| Root | Value |
|---|---|
| Program | 53d73dad…8fc8be59 |
| Weights | 3a9f1e68…6db86f93 |
| Init | 86ecb892…4fbfce1d |
Files
| File | Bytes | SHA-256 |
|---|---|---|
| Program | 4,337,750 | 7de6faf6…b4ff8b30 |
| Weights | 10,064,724 | 3a90fef9…bdc1327d |
Tokenizer (vocab.txt) | 231,508 | 07eced37…992038a3 |
market
Reads a market headline or statement. Answers positive, neutral or negative.
| Source weights | 9mark9/finbert-minilm-sentiment, MIT |
| Architecture | MiniLM in the BERT layout, 12 layers, width 384, 12 heads |
| Parameters | 33,361,155 |
| Input | [CLS] text [SEP] in uncased WordPiece, at most 32 tokens |
| Output | 1 token: the label |
| Steps | 1,318,668 |
| State pages | 1,063 |
| Weight chunks | 133,596 |
The output token is the index of the label:
| Token | Label |
|---|---|
0 | negative |
1 | neutral |
2 | positive |
The SDK exports the same list as MARKET_LABELS. Every API shape that carries either model's output carries the label beside the token.
Asking the market model takes one text:
const { receipt } = await laplace.ask('market', {
text: 'Nvidia beats revenue estimates and raises its outlook',
})A text longer than 30 tokens keeps its first 30, between [CLS] and [SEP]. The headline above is twelve tokens, and the model answers 2, positive.
Roots
| Root | Value |
|---|---|
| Program | ca724dd2…91270cd4 |
| Weights | 8e8edab1…217ff276 |
| Init | 8f8dfa51…6914aa3e |
Files
| File | Bytes | SHA-256 |
|---|---|---|
| Program | 42,222,837 | ecdbe7fc…aca4de3b |
| Weights | 34,903,360 | 4d181dbe…ed568e84 |
Tokenizer (vocab.txt) | 231,508 | 07eced37…992038a3 |
The tokenizer is the judge's file, byte for byte. The program file goes to a client that accepts gzip as 5,634,976 bytes.
Side by side
judge | market | |
|---|---|---|
| Steps | 134,491 | 1,318,668 |
| Turns to one step in a dispute | 6 | 7 |
Largest prove_step payload | 3,351 bytes | 3,511 bytes |
| Program, weights and tokenizer files | 14.6 MB | 77.4 MB |
| A full run in Node | 1.2 s | about 12 s |
A full run is what an operator does to answer and what a watcher or a browser does to run it again. The turns are the most bisect moves a dispute takes to narrow the whole run to one step. The payload is what One step on chain carries when a dispute reaches its last step.
Window and bond
Each model's window and bond are set at registration and never change, and they are read from the network rather than from this page:
- in the app, on
/models/judgeand/models/market, - over HTTP, as
windowSecondsandbondLamportsinGET /v1/models/:slug, - on chain, as
window_secsandbond_lamportsin the model's account at seeds["model", slug].
The same places list the model's live offers, its operators and its recent receipts and disputes.
Fetching a model's files
Every file is served by its SHA-256 from the network's file store, and every client checks the bytes against the hash before it uses them.
| From | With |
|---|---|
| HTTP | GET /v1/models/:slug/files returns the four files with their hashes, sizes and URLs; GET /v1/files/:hash returns one file's bytes, gzip-compressed when the request's Accept-Encoding names gzip. |
| TypeScript | laplace.model(slug) fetches the files, checks each against its hash, loads them into the VM and checks the roots against the model's account. |
| CLI | laplace serve and laplace watch fetch the files of the models named in --models. |
A file whose bytes do not hash to the hash it was fetched by is refused, and the SDK then checks the loaded program and weights against the roots in the model's account.