laion/sera-subset-mixed-3160-axolotl__Qwen3-8B-v8

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:32kPublished:Apr 29, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The laion/sera-subset-mixed-3160-axolotl__Qwen3-8B-v8 is an 8 billion parameter Qwen3-based causal language model, fine-tuned using axolotl on a mixed subset of the `ethanlshen/sera-subset` dataset. This model is specifically optimized for tasks related to the SERA recipe, focusing on agentic reasoning and problem-solving, with a notable context length of 32768 tokens. It is designed for applications requiring robust performance in structured reasoning and agent-like interactions.

Loading preview...

Model Overview

laion/sera-subset-mixed-3160-axolotl__Qwen3-8B-v8 is an 8 billion parameter language model built upon the Qwen3 architecture. It has been instruction fine-tuned (SFT) using the axolotl framework, specifically targeting agentic capabilities.

Key Characteristics

  • Base Model: Qwen3-8B.
  • Training Data: Fine-tuned on a 3160-row random mixed subset of the ethanlshen/sera-subset dataset, which includes both unresolved (stage1) and resolved (stage2) entries.
  • Context Length: Supports a substantial context window of 32768 tokens, enabling processing of longer inputs and complex interactions.
  • Training Recipe: Follows the upstream SERA recipe, indicating an optimization for agent-like reasoning and problem-solving tasks.
  • Hyperparameters: Utilizes a learning rate of 1e-5, a global batch size of 32, 3 epochs, and a sequence length matching its context window. Training was conducted using bf16 precision and DeepSpeed Zero3.

Intended Use Cases

This model is particularly well-suited for applications requiring:

  • Agentic Reasoning: Tasks that benefit from structured problem-solving and multi-step reasoning, as implied by its SERA recipe training.
  • Complex Interaction: Its large context window makes it suitable for handling detailed prompts and maintaining coherence over extended dialogues or problem descriptions.
  • Research and Development: Ideal for further experimentation and development within the domain of agent-based AI systems, especially those leveraging the SERA methodology.

Finetunes
1 models