etiennebamas/qwen3-lr-2-e-5-neat-packing

TEXT GENERATIONPricing:Input $0.468 / Output $1.82Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 27, 2026License:otherArchitecture:Transformer Featherless Exclusive Cold

The etiennebamas/qwen3-lr-2-e-5-neat-packing model is an 8-billion parameter language model, fine-tuned from formalmathatepfl/qwen3-cpt. This model was trained with a learning rate of 2e-05 and a cosine learning rate scheduler over one epoch. It is based on the Qwen3 architecture and has a context length of 32768 tokens.

Loading preview...

Model Overview

This model, etiennebamas/qwen3-lr-2-e-5-neat-packing, is an 8-billion parameter language model. It is a fine-tuned version of the formalmathatepfl/qwen3-cpt model, indicating its foundation in the Qwen3 architecture.

Training Details

The fine-tuning process involved specific hyperparameters:

  • Learning Rate: 2e-05
  • Optimizer: ADAMW_TORCH with betas=(0.9, 0.999) and epsilon=1e-08
  • LR Scheduler: Cosine type with a warmup ratio of 0.05
  • Epochs: 1.0
  • Batch Size: A train_batch_size of 1 and eval_batch_size of 8 were used, leading to a total_train_batch_size of 8 and total_eval_batch_size of 64 across 8 devices.

The model was trained using Transformers 4.57.3, Pytorch 2.9.0+cu128, Datasets 3.6.0, and Tokenizers 0.22.2.

Current Status

As per the provided information, further details regarding the model's description, intended uses, limitations, and training/evaluation data are needed to fully understand its specific capabilities and optimal applications.