A redesigned Billing Page UI has been implemented for improved usability and functionality. Features like Auto Recharge, Change Plan, and the Plan Calculator have been integrated to streamline account management processes.

MonsterAPI now supports Mixtral 8x7B model deployment, enabling hosting Mixtral as an API endpoint on the GPU cloud. Query and fetch responses on-demand with Monster Deploy, the optimized LLM deployment engine designed for higher throughput and lower cost with batching.

Transitioned from overage billing to auto-recharge method for improved service continuity. When the credit balance falls below 5,000, it automatically recharges to maintain a balance above 0 credits, simplifying credit monitoring.

Two new models have been added to our no-code LLM finetuner:

A temporary file size limit of 275 MB is being implemented for our speech-to-text APIs. This limit is sufficient to accommodate audio files of up to 3 hours in length. Users are advised to adjust their API calls accordingly. This adjustment aims to enhance the stability of our speech-to-text transcription systems, resulting in an almost 100% success rate.

MonsterTuner now supports checkpointing, allowing users to terminate a fine-tuning job before completion and still access the last saved checkpoint. Checkpoints are auto-saved periodically during training, providing a way to validate experiments in progress.

A Colab notebook is available for deploying the TinyLlama 1.1B model on MonsterAPI. This model, trained on 1T tokens, offers cost-effective chat application capabilities.

📘

Support for JSON Payload