---

license: gemma
datasets:
- anthracite-org/stheno-filtered-v1.1
base_model: google/gemma-2-2b-it

---

![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXeiuCm7c8lEwEJuRey9kiVZsRn2W-b4pWlu3-X534V3YmVuVc2ZL-NXg2RkzSOOS2JXGHutDuyyNAUtdJI65jGTo8jT9Y99tMi4H4MqL44Uc5QKG77B0d6-JfIkZHFaUA71-RtjyYZWVIhqsNZcx8-OMaA?key=xt3VSDoCbmTY7o-cwwOFwQ)

# QuantFactory/Gemma-2-2B-Stheno-Filtered-GGUF
This is quantized version of [SaisExperiments/Gemma-2-2B-Stheno-Filtered](https://huggingface.co/SaisExperiments/Gemma-2-2B-Stheno-Filtered) created using llama.cpp

# Original Model Card


![image/png](https://cdn-uploads.huggingface.co/production/uploads/660e67afe23148df7ca321a5/F1TQkG-VUmlTFL-xtk3wW.png)

I don't have anything else so you get a cursed cat image

# Basic info
This is [anthracite-org/stheno-filtered-v1.1](https://huggingface.co/datasets/anthracite-org/stheno-filtered-v1.1) over [unsloth/gemma-2-2b-it](https://huggingface.co/unsloth/gemma-2-2b-it)

It saw 76.6M tokens

This time it took 14 hours and i'm pretty sure i've been training with the wrong prompt template X-X

# Training config:
```
cutoff_len: 1024
dataset: stheno-3.4
dataset_dir: data
ddp_timeout: 180000000
do_train: true
finetuning_type: lora
flash_attn: auto
fp16: true
gradient_accumulation_steps: 8
include_num_input_tokens_seen: true
learning_rate: 5.0e-05
logging_steps: 5
lora_alpha: 64
lora_dropout: 0
lora_rank: 64
lora_target: all
lr_scheduler_type: cosine
max_grad_norm: 1.0
max_samples: 100000
model_name_or_path: unsloth/gemma-2-2b-it
num_train_epochs: 3.0
optim: adamw_8bit
output_dir: saves/Gemma-2-2B-Chat/lora/stheno
packing: false
per_device_train_batch_size: 2
plot_loss: true
preprocessing_num_workers: 16
quantization_bit: 4
quantization_method: bitsandbytes
report_to: none
save_steps: 100
stage: sft
template: gemma
use_unsloth: true
warmup_steps: 0
```