A fine-tuned version of unsloth/Qwen3.5-2B trained on Off Grid Survival sharegpt 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 Off Grid Survival sharegpt 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	`data/Off-Grid-Survival-sharegpt.json`
Fine-tuning epochs	2
Fine-tuning date	2026-07-10
Fine-tuning method	LoRA (merged to full 16-bit)

Training Hyperparameters

LoRA

Parameter	Value
`r`	`16`
`alpha`	`64`
`dropout`	`0.06`
`target_modules`	`['q_proj', 'v_proj']`

Training

Parameter	Value
`learning_rate`	`2e-05`
`batch_size`	`4`
`gradient_accumulation_steps`	`4`
`warmup_ratio`	`0.03`
`max_seq_length`	`2048`
`quantization`	`none`

Usage

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model     = AutoModelForCausalLM.from_pretrained("theprint/Survivor-v1-2B")
tokenizer = AutoTokenizer.from_pretrained("theprint/Survivor-v1-2B")