File size: 1,857 Bytes
7d52a38 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 |
---
license: gemma
datasets:
- anthracite-org/stheno-filtered-v1.1
base_model: google/gemma-2-2b-it
---
# 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
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
```
|