Model Description

This is a finetune of StableLM-Zephyr-3B with 2 datasets, toxic-dpo and openhermes with 10000 samples. This was finetuned at 1024 context, for 4k version, go here: https://huggingface.co/Walmart-the-bag/zephyr-quiklang-3b-4K.

Training Parameters

  • 1xA6000-48GB
  • batch_size: 6
  • learning_rate: 5e-5

Datasets:

  • unalignment/toxic-dpo-v0.1
  • teknium/openhermes

Metrics/Basic Eval:

"predict_bleu-4": 31.594154999999997,
"predict_rouge-1": 44.092935,
"predict_rouge-2": 22.276081000000005,
"predict_rouge-l": 34.506909,
"predict_runtime": 121.7549,
"predict_samples_per_second": 0.821,
"predict_steps_per_second": 0.107
Downloads last month
29
Safetensors
Model size
2.8B params
Tensor type
F32
·
FP16
·
Inference Examples
Inference API (serverless) has been turned off for this model.

Model tree for Walmart-the-bag/zephyr-quiklang-3b

Finetuned
(11)
this model
Finetunes
1 model
Quantizations
2 models

Datasets used to train Walmart-the-bag/zephyr-quiklang-3b

Collection including Walmart-the-bag/zephyr-quiklang-3b