runbios-sdk 0.2.5 → 0.2.6-dev.198

package/README.md

@@ -248,6 +248,12 @@ const imported = await client.datasets.importFromHuggingFace({
 
 // Validate before upload
 const validation = await client.datasets.validate('./data.jsonl');
+
+// Maintain an existing dataset explicitly
+await client.datasets.updateColumnMapping('ds_abc123', { conversation: 'messages' });
+await client.datasets.revalidate('ds_abc123');
+await client.datasets.updateSource('ds_abc123'); // reference-mode Hub datasets only
+await client.datasets.setIntegration('ds_abc123', 'int_replacement'); // null detaches
 ```
 
 #### Connected Hub search and exact source selection
@@ -273,9 +279,14 @@ polling timeout does not cancel the background import.
 
 ### Training
 
-`datasetIds` preserves source order. Pass the same `mixing` object to preflight
-and create: `{ mode: 'interleave', weights: [3, 1], seed: 42 }`. Weights match the
-ordered IDs. Modes are `sequential`, `shuffle`, `interleave`, and `phased`; phased
+`datasetIds` preserves source order. `evalDatasetIds` selects separate held-out
+datasets and is mutually exclusive with `evalSplit`; evaluation datasets are
+composed sequentially and are never optimization rows. Both sets are frozen at
+job creation—launch a new job to change them; an active run cannot accept more
+data. Pass the same `mixing`
+object to preflight and create: `{ mode: 'interleave', weights: [3, 1], seed: 42 }`.
+Weights match the ordered training IDs. Modes are `sequential`, `shuffle`,
+`interleave`, and `phased`; phased
 plans contain ordered `phases` with `name`, `portion`, optional `weights`, and
 `shuffle`. The default composition seed is 42. Resume reconstructs data from
 source pins and the stored plan and verifies the checksum.
@@ -285,6 +296,7 @@ const request = {
   idempotencyKey: 'training-create-20260711-0001',
   model: 'meta-llama/Llama-3.1-8B-Instruct',
   datasetIds: ['ds_abc123', 'ds_def456'],
+  evalDatasetIds: ['ds_holdout'],
   method: 'sft' as const,
   adapter: 'lora' as const,
   queueIfUnavailable: true,

package/dist/client.js

@@ -340,7 +340,7 @@ export class HttpClient {
     constructor(config) {
         // Default host stays api.runbios.ai for now; cutover to api.runbios.ai is
         // planned once its DNS exists.
-        this.baseUrl = (config.baseUrl || envBaseUrl() || 'https://api.runbios.ai').replace(/\/+$/, '');
+        this.baseUrl = (config.baseUrl || envBaseUrl() || 'https://api-dev.runbios.ai').replace(/\/+$/, '');
         this.apiKey = config.apiKey ?? envApiKey();
         this.accessToken = config.accessToken;
         this.orgId = config.orgId;

package/dist/index.d.ts

@@ -36,7 +36,7 @@ import { Loop } from './resources/loop.js';
  * SDK version. Sent as part of the User-Agent header.
  * Must match package.json "version" -- enforced by a contract test.
  */
-export declare const VERSION = "0.2.5";
+export declare const VERSION = "0.2.6-dev.198";
 export declare class RunBiOS {
     /** Search models, fetch configs, check adapter compatibility. */
     readonly models: Models;

package/dist/index.js

@@ -36,7 +36,7 @@ import { Loop } from './resources/loop.js';
  * SDK version. Sent as part of the User-Agent header.
  * Must match package.json "version" -- enforced by a contract test.
  */
-export const VERSION = '0.2.5';
+export const VERSION = '0.2.6-dev.198';
 export class RunBiOS {
     /** Search models, fetch configs, check adapter compatibility. */
     models;

package/dist/resources/datasets.d.ts

@@ -1,5 +1,5 @@
 import type { HttpClient } from '../client.js';
-import type { Dataset, DatasetListParams, DatasetListResponse, DatasetUploadParams, DatasetPreview, DatasetPreviewParams, DatasetImportHFParams, DatasetRegisterHFParams, DatasetHubSearchParams, DatasetHubPreviewParams, DatasetValidation, DatasetFormatSpecs, DatasetStorageUsage } from '../types.js';
+import type { Dataset, DatasetListParams, DatasetListResponse, DatasetUploadParams, DatasetPreview, DatasetPreviewParams, DatasetImportHFParams, DatasetRegisterHFParams, DatasetRegisterLoopParams, DatasetHubSearchParams, DatasetHubPreviewParams, DatasetValidation, DatasetFormatSpecs, DatasetStorageUsage } from '../types.js';
 /**
  * Manage training datasets -- upload files, import from HuggingFace,
  * preview contents, and manage storage.
@@ -76,6 +76,14 @@ export declare class Datasets {
      * ```
      */
     delete(id: string): Promise<void>;
+    /** Replace the source mapping and immediately revalidate the dataset. */
+    updateColumnMapping(id: string, columnMapping: Record<string, string>): Promise<Record<string, unknown>>;
+    /** Revalidate an existing dataset against its pinned source. */
+    revalidate(id: string): Promise<Record<string, unknown>>;
+    /** Adopt a compatible newer revision of a reference-mode Hub dataset. */
+    updateSource(id: string): Promise<Record<string, unknown>>;
+    /** Attach another connected Hub integration, or detach it with null. */
+    setIntegration(id: string, integrationId: string | null): Promise<Record<string, unknown>>;
     /**
      * Import a dataset from HuggingFace Hub via a connected integration.
      *
@@ -110,6 +118,30 @@ export declare class Datasets {
      * ```
      */
     registerHuggingFace(params: DatasetRegisterHFParams | Record<string, unknown>): Promise<Dataset>;
+    /**
+     * Turn a curated Conscious Loop set into a dataset training can use.
+     *
+     * Without this, using a loop set from code meant downloading the JSONL and
+     * uploading it again as an unrelated dataset. That works, and it throws away
+     * the one thing that made the set worth having: which conversations, which
+     * verdicts and which judges produced it. Registering keeps that link.
+     *
+     * The set is then put through the SAME validation every uploaded dataset
+     * gets, so it comes back `processing` rather than ready -- poll
+     * {@link getStatus} before training on it. Registering the same set and split
+     * twice is refused rather than billed twice, and an empty set is refused
+     * rather than registered as something that would fail inside the trainer.
+     *
+     * @example
+     * ```ts
+     * const train = await client.datasets.registerLoop({
+     *   loopDatasetId: 'd244894c-7ab4-4b3d-bfda-e0bb4c563ceb',
+     *   name: 'extract-v1 distill SFT',
+     *   split: 'train',
+     * });
+     * ```
+     */
+    registerLoop(params: DatasetRegisterLoopParams | Record<string, unknown>): Promise<Dataset>;
     /**
      * Get the processing status of a dataset.
      *

package/dist/resources/datasets.js

@@ -173,6 +173,22 @@ export class Datasets {
     async delete(id) {
         await this._http.fetchDelete(`/api/datasets/${encodeURIComponent(id)}`);
     }
+    /** Replace the source mapping and immediately revalidate the dataset. */
+    async updateColumnMapping(id, columnMapping) {
+        return this._http.fetchPatch(`/api/datasets/${encodeURIComponent(id)}/column-mapping`, columnMapping);
+    }
+    /** Revalidate an existing dataset against its pinned source. */
+    async revalidate(id) {
+        return this._http.fetchPost(`/api/datasets/${encodeURIComponent(id)}/revalidate`, {});
+    }
+    /** Adopt a compatible newer revision of a reference-mode Hub dataset. */
+    async updateSource(id) {
+        return this._http.fetchPost(`/api/datasets/${encodeURIComponent(id)}/update-source`, {});
+    }
+    /** Attach another connected Hub integration, or detach it with null. */
+    async setIntegration(id, integrationId) {
+        return this._http.fetchPost(`/api/datasets/${encodeURIComponent(id)}/integration`, { integration_id: integrationId });
+    }
     /**
      * Import a dataset from HuggingFace Hub via a connected integration.
      *
@@ -244,6 +260,45 @@ export class Datasets {
         });
         return normalizeDataset(result);
     }
+    /**
+     * Turn a curated Conscious Loop set into a dataset training can use.
+     *
+     * Without this, using a loop set from code meant downloading the JSONL and
+     * uploading it again as an unrelated dataset. That works, and it throws away
+     * the one thing that made the set worth having: which conversations, which
+     * verdicts and which judges produced it. Registering keeps that link.
+     *
+     * The set is then put through the SAME validation every uploaded dataset
+     * gets, so it comes back `processing` rather than ready -- poll
+     * {@link getStatus} before training on it. Registering the same set and split
+     * twice is refused rather than billed twice, and an empty set is refused
+     * rather than registered as something that would fail inside the trainer.
+     *
+     * @example
+     * ```ts
+     * const train = await client.datasets.registerLoop({
+     *   loopDatasetId: 'd244894c-7ab4-4b3d-bfda-e0bb4c563ceb',
+     *   name: 'extract-v1 distill SFT',
+     *   split: 'train',
+     * });
+     * ```
+     */
+    async registerLoop(params) {
+        const raw = params;
+        const loopDatasetId = String(raw.loopDatasetId || raw.loop_dataset_id || '');
+        if (!loopDatasetId)
+            throw new Error('RunBiOS: loopDatasetId is required');
+        const workspaceId = String(raw.workspaceId || raw.workspace_id || '')
+            || this._http.workspaceId
+            || await this._http.resolveWorkspaceId();
+        const result = await this._http.fetchPost('/api/datasets/register-loop', {
+            loop_dataset_id: loopDatasetId,
+            workspace_id: workspaceId,
+            split: raw.split ?? 'all',
+            ...(raw.name !== undefined && { name: raw.name }),
+        });
+        return normalizeDataset(result);
+    }
     /**
      * Get the processing status of a dataset.
      *

package/dist/resources/inference.js

@@ -326,7 +326,7 @@ export class Inference {
         this.key = config.inferenceKey || envInferenceKey() || envApiKey();
         // Same default host as HttpClient (api.runbios.ai); api.runbios.ai cutover
         // is planned once its DNS exists — update both call sites together.
-        this.baseUrl = (config.baseUrl || envBaseUrl() || 'https://api.runbios.ai').replace(/\/+$/, '');
+        this.baseUrl = (config.baseUrl || envBaseUrl() || 'https://api-dev.runbios.ai').replace(/\/+$/, '');
         this.timeout = config.timeout ?? 900_000;
         this._http = http;
     }

package/dist/resources/training.js

@@ -19,12 +19,15 @@ function buildTrainingRequest(params) {
     if (datasetIds.length === 0) {
         throw new Error('RunBiOS: datasetId or datasetIds must contain at least one dataset');
     }
+    const evalDatasetIds = [...new Set((params.evalDatasetIds || []).map((id) => id.trim()).filter(Boolean))];
     const body = {
         model_id: params.model,
         dataset_ids: datasetIds,
         training_method: trainingMethod,
         train_type: params.adapter,
     };
+    if (evalDatasetIds.length > 0)
+        body.eval_dataset_ids = evalDatasetIds;
     if (params.modelRevision !== undefined)
         body.model_revision = params.modelRevision;
     const queueEnabled = params.queueIfUnavailable ?? false;
@@ -238,6 +241,7 @@ export class Training {
             training_method: trainingMethod,
             train_type: params.adapter,
             dataset_ids: datasetIds,
+            eval_dataset_ids: [...new Set((params.evalDatasetIds || []).map((id) => id.trim()).filter(Boolean))],
             gpu_type: params.gpuType,
             gpu_count: params.gpuCount,
             status: 'pending',

package/dist/types.d.ts

@@ -8,13 +8,13 @@ export interface BiOSConfig {
     orgId?: string;
     /** Workspace ID. Optional when using API keys (resolved from the key). Can override for multi-workspace keys. */
     workspaceId?: string;
-    /** Base URL for the API. Falls back to the RUNBIOS_BASE_URL environment variable (legacy: BIOS_BASE_URL), then the canonical https://api.runbios.ai hostname. */
+    /** Base URL for the API. Falls back to the RUNBIOS_BASE_URL environment variable (legacy: BIOS_BASE_URL), then the canonical https://api-dev.runbios.ai hostname. */
     baseUrl?: string;
     /** Request timeout in milliseconds. Defaults to 30000. */
     timeout?: number;
     /** Default per-deployment inference key. Can be overridden per inference call. */
     inferenceKey?: string;
-    /** Inference base URL. Defaults to baseUrl, then https://api.runbios.ai. */
+    /** Inference base URL. Defaults to baseUrl, then https://api-dev.runbios.ai. */
     inferenceBaseUrl?: string;
     /** End-to-end inference timeout in milliseconds. Defaults to 15 minutes. */
     inferenceTimeout?: number;
@@ -286,12 +286,26 @@ export interface ModelConfig {
     /** Whether config was successfully resolved. */
     resolved: boolean;
 }
+export interface DatasetCatalog {
+    slug: string;
+    version: number;
+    access_mode: 'everyone' | 'organizations' | 'tiers';
+    license: string;
+    attribution: string;
+    source_url: string;
+    source_revision: string;
+}
 /** A dataset in the user's workspace. */
 export interface Dataset {
     id: string;
     name: string;
     description?: string;
-    workspace_id?: string;
+    workspace_id?: string | null;
+    is_system?: boolean;
+    provider?: 'RunBIOS' | null;
+    storage_billing_owner?: 'runbios' | 'customer';
+    read_only?: boolean;
+    catalog?: DatasetCatalog | null;
     source?: string;
     hf_dataset_id?: string | null;
     hf_subset?: string | null;
@@ -430,6 +444,21 @@ export interface DatasetRegisterHFParams {
     columnMapping?: Record<string, string>;
     importMode?: 'auto' | 'reference' | 'materialize';
 }
+/** Parameters for registering a curated Conscious Loop set as a dataset. */
+export interface DatasetRegisterLoopParams {
+    /** The Conscious Loop dataset to register. */
+    loopDatasetId: string;
+    /** Defaults to `loop-<first 8 of the loop id>`. */
+    name?: string;
+    /**
+     * Which rows to take. Defaults to `all`, because the loop already split the
+     * set by TIME and re-splitting a deliberate split throws that decision away.
+     * Register `train` and `holdout` separately to score a training job against
+     * the loop's own held-back rows; both can exist at once.
+     */
+    split?: 'all' | 'train' | 'holdout';
+    workspaceId?: string;
+}
 /** Parameters for searching HuggingFace Hub datasets. */
 export interface DatasetHubSearchParams {
     /** Search query. */
@@ -541,6 +570,8 @@ export interface TrainingCreateParams {
     datasetId?: string;
     /** Ordered dataset IDs for multi-dataset training. */
     datasetIds?: string[];
+    /** Separate evaluation datasets, composed sequentially. Mutually exclusive with evalSplit. */
+    evalDatasetIds?: string[];
     /**
      * Optional per-dataset row cap: keep only the first N rows of a dataset
      * (key = a dataset ID from datasetIds, value = row count). Omit an entry, or
@@ -695,6 +726,7 @@ export interface TrainingJob {
     train_type?: AdapterType;
     rlhf_type?: RLHFAlgorithm | null;
     dataset_ids?: string[];
+    eval_dataset_ids?: string[];
     gpu_type?: string;
     gpu_count?: number;
     storage_gb?: number;
@@ -889,6 +921,7 @@ export interface CanonicalTrainingRequest {
     name: string;
     dataset_id: string;
     dataset_ids: string[];
+    eval_dataset_ids?: string[];
     hf_dataset_ids?: string;
     dataset_mixing?: string;
     mixing?: Record<string, unknown>;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "runbios-sdk",
-  "version": "0.2.5",
+  "version": "0.2.6-dev.198",
   "description": "Official TypeScript SDK for the Run BiOS training and deployment platform API",
   "type": "module",
   "main": "./dist/index.js",