runbios-sdk 0.2.1-dev.100 → 0.2.1-dev.104

package/dist/index.d.ts

@@ -34,7 +34,7 @@ import { Inference } from './resources/inference.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 = "0.2.1-dev.100";
+export declare const VERSION = "0.2.1-dev.104";
 export declare class RunBiOS {
     /** Search models, fetch configs, check adapter compatibility. */
     readonly models: Models;

package/dist/index.js

@@ -34,7 +34,7 @@ import { Inference } from './resources/inference.js';
  * SDK version. Sent as part of the User-Agent header.
  * Must match package.json "version" -- enforced by a contract test.
  */
-export const VERSION = '0.2.1-dev.100';
+export const VERSION = '0.2.1-dev.104';
 export class RunBiOS {
     /** Search models, fetch configs, check adapter compatibility. */
     models;

package/dist/resources/datasets.d.ts

@@ -84,10 +84,14 @@ export declare class Datasets {
      *
      * @example
      * ```ts
+     * The selected split must already expose canonical rows: a `messages` array
+     * for SFT or a `text` field for CPT. Legacy layouts are rejected with
+     * conversion guidance rather than remapped silently.
+     *
      * const imported = await client.datasets.importFromHuggingFace({
-     *   repoId: 'databricks/dolly-15k',
+     *   repoId: 'your-org/dataset-with-messages',
      *   integrationId: 'int_abc123',
-     *   name: 'Dolly 15k',
+     *   name: 'Canonical SFT data',
      * });
      * console.log(`Imported: ${imported.id}`);
      * ```
@@ -99,8 +103,8 @@ export declare class Datasets {
      * @example
      * ```ts
      * const ds = await client.datasets.registerHuggingFace({
-     *   repo_id: 'databricks/dolly-15k',
-     *   name: 'Dolly 15k',
+     *   repo_id: 'your-org/dataset-with-text',
+     *   name: 'Canonical CPT corpus',
      *   split: 'train',
      * });
      * ```

package/dist/resources/datasets.js

@@ -114,6 +114,9 @@ export class Datasets {
             const message = err instanceof Error ? err.message : String(err);
             throw new Error(`Failed to read file at "${params.filePath}": ${message}`);
         }
+        if (!fileName.toLowerCase().endsWith('.jsonl')) {
+            throw new Error('RunBiOS: dataset uploads accept .jsonl only — one JSON object per line, with "messages" for SFT or "text" for CPT');
+        }
         const blob = new Blob([fileData]);
         const formData = new FormData();
         formData.append('file', blob, fileName);
@@ -181,10 +184,14 @@ export class Datasets {
      *
      * @example
      * ```ts
+     * The selected split must already expose canonical rows: a `messages` array
+     * for SFT or a `text` field for CPT. Legacy layouts are rejected with
+     * conversion guidance rather than remapped silently.
+     *
      * const imported = await client.datasets.importFromHuggingFace({
-     *   repoId: 'databricks/dolly-15k',
+     *   repoId: 'your-org/dataset-with-messages',
      *   integrationId: 'int_abc123',
-     *   name: 'Dolly 15k',
+     *   name: 'Canonical SFT data',
      * });
      * console.log(`Imported: ${imported.id}`);
      * ```
@@ -210,8 +217,8 @@ export class Datasets {
      * @example
      * ```ts
      * const ds = await client.datasets.registerHuggingFace({
-     *   repo_id: 'databricks/dolly-15k',
-     *   name: 'Dolly 15k',
+     *   repo_id: 'your-org/dataset-with-text',
+     *   name: 'Canonical CPT corpus',
      *   split: 'train',
      * });
      * ```

package/dist/resources/training.d.ts

@@ -12,8 +12,11 @@ export declare class Training {
     /**
      * Create a new training job (book-before-reveal).
      *
-     * The call blocks while the ranked GPU ladder is booked (~40s typical). A
-     * training id and the "training started" email exist only once a real machine is
+     * Choose method, adapter, datasets and the complete hyperparameter config
+     * first, call {@link preflight}, then choose one returned GPU configuration
+     * last. Its minimum is derived from every earlier choice. This call blocks
+     * while that placement is booked (~40s typical). A training id and the
+     * "training started" email exist only once a real machine is
      * secured, so the returned `status` is one of:
      *
      * - `"booked"` — a GPU was secured (booked == secured); the job then
@@ -25,11 +28,12 @@ export declare class Training {
      *   (`queueIfUnavailable: true`); waits for stock at zero charge and books
      *   via the same path.
      *
-     * If the ladder is exhausted at booking time without queue consent, this
-     * rejects with a `CAPACITY_UNAVAILABLE` (HTTP 409) carrying neutral
-     * alternatives — no phantom job remains and nothing is billed. The SDK never
-     * auto-substitutes a GPU; a transient 503 is a retry, never a capacity
-     * verdict.
+     * If booking fails, this rejects with `CAPACITY_UNAVAILABLE` (HTTP 409)
+     * carrying fresh qualifying alternatives — no phantom job and no charge.
+     * Keep every earlier field unchanged, select one alternative and retry. The
+     * rejected type is never recommended back. Queue consent is appropriate only
+     * when the alternatives list is empty; waiting is unbilled. The SDK never
+     * auto-substitutes a GPU; a transient 503 is a retry, never a capacity verdict.
      *
      * @example
      * ```ts

package/dist/resources/training.js

@@ -67,8 +67,6 @@ function buildTrainingRequest(params) {
         body.integration_id = params.integrationId;
     if (params.networkVolumeId !== undefined)
         body.network_volume_id = params.networkVolumeId;
-    if (params.cacheDataset !== undefined)
-        body.cache_dataset = params.cacheDataset;
     if (params.datasetSampleLimits !== undefined)
         body.dataset_sample_limits = params.datasetSampleLimits;
     if (params.datasetMixing !== undefined)
@@ -78,6 +76,8 @@ function buildTrainingRequest(params) {
     const config = {};
     if (params.epochs !== undefined)
         config.num_train_epochs = params.epochs;
+    if (params.maxSteps !== undefined)
+        config.max_steps = params.maxSteps;
     if (params.batchSize !== undefined)
         config.per_device_train_batch_size = params.batchSize;
     if (params.gradientAccumulation !== undefined)
@@ -183,8 +183,11 @@ export class Training {
     /**
      * Create a new training job (book-before-reveal).
      *
-     * The call blocks while the ranked GPU ladder is booked (~40s typical). A
-     * training id and the "training started" email exist only once a real machine is
+     * Choose method, adapter, datasets and the complete hyperparameter config
+     * first, call {@link preflight}, then choose one returned GPU configuration
+     * last. Its minimum is derived from every earlier choice. This call blocks
+     * while that placement is booked (~40s typical). A training id and the
+     * "training started" email exist only once a real machine is
      * secured, so the returned `status` is one of:
      *
      * - `"booked"` — a GPU was secured (booked == secured); the job then
@@ -196,11 +199,12 @@ export class Training {
      *   (`queueIfUnavailable: true`); waits for stock at zero charge and books
      *   via the same path.
      *
-     * If the ladder is exhausted at booking time without queue consent, this
-     * rejects with a `CAPACITY_UNAVAILABLE` (HTTP 409) carrying neutral
-     * alternatives — no phantom job remains and nothing is billed. The SDK never
-     * auto-substitutes a GPU; a transient 503 is a retry, never a capacity
-     * verdict.
+     * If booking fails, this rejects with `CAPACITY_UNAVAILABLE` (HTTP 409)
+     * carrying fresh qualifying alternatives — no phantom job and no charge.
+     * Keep every earlier field unchanged, select one alternative and retry. The
+     * rejected type is never recommended back. Queue consent is appropriate only
+     * when the alternatives list is empty; waiting is unbilled. The SDK never
+     * auto-substitutes a GPU; a transient 503 is a retry, never a capacity verdict.
      *
      * @example
      * ```ts

package/dist/types.d.ts

@@ -565,8 +565,10 @@ export interface TrainingCreateParams {
     name?: string;
     /** Target workspace ID. */
     workspaceId?: string;
-    /** Number of training epochs. */
+    /** Number of training epochs. Governs the full run when maxSteps is omitted. */
     epochs?: number;
+    /** Optional hard step cap that overrides epochs. Omit for no cap and all configured epochs. */
+    maxSteps?: number;
     /** Training batch size per device. */
     batchSize?: number;
     /** Gradient accumulation steps. */
@@ -609,7 +611,7 @@ export interface TrainingCreateParams {
     integrationId?: string;
     /** Existing network volume to attach. */
     networkVolumeId?: string;
-    /** Cache the composed dataset for retry/resume. Defaults to true server-side. */
+    /** @deprecated No effect. Prepared data is transient; exact resume rebuilds it from pinned source metadata and verifies its checksum. */
     cacheDataset?: boolean;
     /** Legacy dataset ordering mode. Prefer mixing for weighted/phased plans. */
     datasetMixing?: 'shuffle' | 'sequential' | 'interleave' | 'random' | 'curriculum';

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "runbios-sdk",
-  "version": "0.2.1-dev.100",
+  "version": "0.2.1-dev.104",
   "description": "Official TypeScript SDK for the Run BiOS training and deployment platform API",
   "type": "module",
   "main": "./dist/index.js",