generated: '2026-08-02'
method: generated
source: openapi/tensorwave-scalarlm-openapi.yml
provider: tensorwave
api: tensorwave:scalarlm
note: 'No provider-published skills or AGENTS.md were found in the TensorWave documentation or in
  any of the 19 repositories in the tensorwavecloud GitHub organization. These are generated by
  API Evangelist and every step is grounded in an operationId that exists verbatim in
  openapi/tensorwave-scalarlm-openapi.yml.'
skills:
- file: tensorwave-openai-compatible-chat.md
  name: Call ScalarLM as an OpenAI-compatible endpoint
  description: Point an existing OpenAI client at a ScalarLM deployment for chat and text
    completions, and handle its two different response transports.
  api: openapi/tensorwave-scalarlm-openapi.yml
  operations: [listModels, createChatCompletion, createCompletion]
- file: tensorwave-batch-inference.md
  name: Run batch inference on a ScalarLM deployment
  description: Submit a batch of prompts to the inference queue, poll for results, and handle
    per-item failures that ride on a 200 response.
  api: openapi/tensorwave-scalarlm-openapi.yml
  operations: [listModels, generate, getResults, listRequests, getRequestDetail,
    getGenerateMetrics]
- file: tensorwave-fine-tune-and-publish.md
  name: Fine-tune a model on ScalarLM and publish it to Hugging Face
  description: Upload a dataset in resumable chunks, launch and monitor a Megatron-LM training job
    on Slurm, then push the resulting checkpoint to the Hugging Face Hub.
  api: openapi/tensorwave-scalarlm-openapi.yml
  operations: [initChunkedUpload, uploadChunk, finalizeChunkedUpload, submitTrainingJob,
    getTrainingJob, getTrainingLogs, getTrainingDataset, listCheckpoints, publishToHuggingFace,
    getPublishStatus, getPublishLogs, cancelPublish, getMegatronSqueue, getGpuCount, getNodeCount]