LinjunChen123/Protoss-a

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:2BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 19, 2026Architecture:Transformer Featherless Exclusive Cold

LinjunChen123/Protoss-a is a 2 billion parameter causal language model, fine-tuned from the Qwen3-1.7B base model using H2O LLM Studio. This model is designed for general text generation tasks, leveraging its Qwen3 architecture and a 32768-token context length. It is suitable for applications requiring conversational AI and text completion, built upon a robust transformer framework.

Loading preview...

Model Overview

LinjunChen123/Protoss-a is a 2 billion parameter causal language model, developed by fine-tuning the Qwen/Qwen3-1.7B base model. The training process utilized H2O LLM Studio, a platform for developing large language models. This model features a Qwen3ForCausalLM architecture, including 28 decoder layers and a 32768-token context length, making it capable of handling substantial input sequences.

Key Capabilities

  • General Text Generation: Capable of generating human-like text based on provided prompts.
  • Conversational AI: Designed to engage in multi-turn conversations, as demonstrated by its usage examples.
  • Flexible Deployment: Supports loading with transformers library, including options for torch_dtype="auto" and device_map for GPU utilization.
  • Quantization Support: Can be loaded in 8-bit or 4-bit quantization for reduced memory footprint, and supports sharding across multiple GPUs.

Good For

  • Developers: Those looking for a Qwen3-based model for text generation and conversational applications.
  • Experimentation: Users interested in exploring models fine-tuned with H2O LLM Studio.
  • Resource-Efficient Deployment: Suitable for environments where quantization and multi-GPU sharding are beneficial for performance and memory management.