DCAgent/staqc-sandboxes-traces-terminus-2_Qwen3-32B

TEXT GENERATIONConcurrency Cost:2Model Size:32BQuant:FP8Ctx Length:32kPublished:Oct 25, 2025License:apache-2.0Architecture:Transformer Open Weights Cold

DCAgent/staqc-sandboxes-traces-terminus-2_Qwen3-32B is a 32 billion parameter causal language model, fine-tuned from the Qwen/Qwen3-32B architecture. This model is specifically adapted using the mlfoundations-dev/staqc-sandboxes-traces-terminus-2 dataset. It is optimized for tasks related to the specific data distribution of its fine-tuning dataset, making it suitable for applications requiring specialized knowledge from that domain.

Loading preview...

Model Overview

This model, staqc-sandboxes-traces-terminus-2_Qwen3-32B, is a fine-tuned variant of the Qwen3-32B base model developed by Qwen. It has been specifically adapted through further training on the mlfoundations-dev/staqc-sandboxes-traces-terminus-2 dataset.

Key Characteristics

  • Base Model: Qwen/Qwen3-32B
  • Parameter Count: 32 billion parameters
  • Context Length: 32768 tokens
  • Fine-tuning Dataset: mlfoundations-dev/staqc-sandboxes-traces-terminus-2

Training Details

The fine-tuning process utilized the following key hyperparameters:

  • Learning Rate: 4e-05
  • Optimizer: ADAMW_TORCH_FUSED
  • Batch Size: A total training batch size of 64 (with 1 per device and 4 gradient accumulation steps) across 16 GPUs.
  • Epochs: 5.0 epochs
  • Scheduler: Cosine learning rate scheduler with a 0.1 warmup ratio.

Intended Use Cases

Given its fine-tuning on a specific dataset, this model is best suited for applications that align with the data distribution and tasks present in the mlfoundations-dev/staqc-sandboxes-traces-terminus-2 dataset. Developers should consider its specialized training for tasks requiring nuanced understanding or generation within that domain.