Edit model card

簡介

本模型是基於MediaTek-Research/Breeze-7B-Instruct-v1_0微調後的產物
模型使用來自PTT網站中的Gossiping分類的資料訓練
過程中使用了一些方法從海量的數據中,過濾出噪聲較小(理論上)的部份作為訓練數據
訓練資料: Riyuechang/PTT-Corpus-100K_Gossiping-1400-39400

設備

  • Ubuntu 22.04.4 LTS
  • NVIDIA GeForce RTX 3060 12G

Lora參數

r=8,
lora_alpha=32,
lora_dropout=0.1,
task_type="CAUSAL_LM",
target_modules="all-linear",
bias="none",
use_dora=True,
use_rslora=True

訓練參數

per_device_train_batch_size=28,  
gradient_accumulation_steps=1,  
num_train_epochs=3,  
warmup_ratio=0.1,  
learning_rate=2e-5,  
bf16=True,  
save_strategy="steps",  
save_steps=500,  
save_total_limit=10,  
logging_steps=10,  
output_dir=log_output,  
optim="paged_adamw_8bit",  
gradient_checkpointing=True

結果

  • loss: 1.1035
Downloads last month
21
Safetensors
Model size
7.49B params
Tensor type
BF16
·
Inference Examples
Unable to determine this model's library. Check the docs .

Model tree for Riyuechang/Breeze-7B-PTT-Chat-v1

Finetuned
(7)
this model
Quantizations
1 model

Dataset used to train Riyuechang/Breeze-7B-PTT-Chat-v1