File size: 821 Bytes
e70d9bf
 
175a00b
 
e70d9bf
b88c7b6
175a00b
 
 
e09bb18
 
 
175a00b
e09bb18
175a00b
 
 
 
 
 
 
 
6cb95b1
175a00b
e09bb18
175a00b
 
 
 
e09bb18
 
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
---
license: mit
datasets:
- yahma/alpaca-cleaned
---
This repo contains a low-rank adapter for LLaMA-13b fit on the Cleaned Alpaca dataset containing the new GPT-4 data.

This version of the weights was trained with the following hyperparameters:

    Cleaned dataset: Snapshot April 9, 2023
    Epochs: 4
    Validation set size: 1500
    Batch size: 128
    Micro batch size: 4
    Cutoff length: 512
    Learning rate: 3e-4
    Lora r: 16
    Lora target modules: q_proj, k_proj, v_proj, o_proj

That is:

python finetune.py \
    --base_model='yahma/llama-13b-hf' \
    --data_path 'yahma/alpaca-cleaned' \
    --num_epochs=4 \
    --cutoff_len=512 \
    --output_dir='./lora-alpaca' \
    --lora_target_modules='[q_proj,k_proj, v_proj, o_proj]' \
    --lora_r=16 \
    --val_set_size 1500 \
    --micro_batch_size=4