gpt-oss-20b-ck-MXFP4

This is a test fine-tune of the base model unsloth/gpt-oss-20b. It was fine-tuned using Unsloth for efficiency on a small dataset as an experimental setup.

Model Details

  • Base Model: unsloth/gpt-oss-20b (MXFP4 quantized)
  • Fine-Tuning Method: QLoRA with rank=64, targeting MoE layers
  • Training Epochs: 6
  • Dataset: Small custom dataset (~4,000 examples)
  • Max Sequence Length: 8192
  • Optimizer: AdamW 8-bit
  • Learning Rate: 1e-4

The model is provided in MXFP4 GGUF format for compatibility with llama.cpp, Ollama, or LM Studio.

Usage

Load with Unsloth or transformers for inference:

from unsloth import FastLanguageModel
model, tokenizer = FastLanguageModel.from_pretrained("Entz/gpt-oss-20b-ck-MXFP4")
Downloads last month
221
GGUF
Model size
21B params
Architecture
gpt-oss
Hardware compatibility
Log In to view the estimation

We're not able to determine the quantization variants.

Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for Entz/gpt-oss-20b-ck-MXFP4

Base model

openai/gpt-oss-20b
Quantized
(17)
this model