A fine-tuned version of unsloth/Qwen3.5-2B trained on theprint Alpaca Docs n Summaries data using Auto-SFT — an automated hyperparameter search and supervised fine-tuning pipeline.

The base model was adapted to follow the style and content of the theprint Alpaca Docs n Summaries dataset. Expect improved performance on tasks similar to those represented in the training data.

Model Details

Property	Value
Base model	`unsloth/Qwen3.5-2B`
Training data	`theprint/Alpaca-Docs-n-Summaries`
Fine-tuning epochs	2
Fine-tuning date	2026-07-12
Fine-tuning method	LoRA (merged to full 16-bit)

Training Hyperparameters

LoRA

Parameter	Value
`r`	`64`
`alpha`	`64`
`dropout`	`0.0`
`target_modules`	`['q_proj', 'v_proj', 'k_proj', 'o_proj', 'gate_proj', 'up_proj', 'down_proj']`

Training

Parameter	Value
`learning_rate`	`1e-05`
`batch_size`	`4`
`gradient_accumulation_steps`	`1`
`warmup_ratio`	`0.05`
`max_seq_length`	`2048`
`quantization`	`none`

GGUF Files

These quantized GGUF files can be used directly with llama.cpp, Ollama, LM Studio, and other compatible runtimes.

File	Description
`Summarizer-v1-2B-GGUF-BF16.gguf`	BF16
`Summarizer-v1-2B-GGUF-Q8_0.gguf`	8-bit — near-lossless, larger file
`Summarizer-v1-2B-GGUF-Q6_K.gguf`	6-bit — high quality
`Summarizer-v1-2B-GGUF-Q5_K_M.gguf`	5-bit medium — good quality/size balance
`Summarizer-v1-2B-GGUF-Q5_K_S.gguf`	Q5_K_S
`Summarizer-v1-2B-GGUF-Q4_K_M.gguf`	4-bit medium — recommended for most use cases
`Summarizer-v1-2B-GGUF-Q4_K_S.gguf`	Q4_K_S
`Summarizer-v1-2B-GGUF-Q3_K_L.gguf`	Q3_K_L
`Summarizer-v1-2B-GGUF-Q3_K_M.gguf`	Q3_K_M
`Summarizer-v1-2B-GGUF-Q3_K_S.gguf`	Q3_K_S
`Summarizer-v1-2B-GGUF-Q2_K.gguf`	2-bit — smallest size, lowest quality
`Summarizer-v1-2B-GGUF-IQ4_NL.gguf`	IQ4_NL

Generated by Auto-SFT

theprint/Summarizer-v1-2B-GGUF

Summarizer-v1-2B (GGUF)

Model Details

Training Hyperparameters

LoRA

Training

GGUF Files

theprint/
Summarizer-v1-2B-GGUF