File size: 5,571 Bytes
9218ebe 03e5907 9218ebe 12a2dbb 9218ebe 12a2dbb 9218ebe 12a2dbb 03e5907 9218ebe 2414673 03e5907 2414673 03e5907 2414673 12a2dbb 03e5907 2414673 9218ebe 03e5907 9218ebe 03e5907 |
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 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 |
"""
E2E tests for lora llama
"""
import logging
import os
import tempfile
import unittest
from pathlib import Path
from axolotl.cli import load_datasets
from axolotl.common.cli import TrainerCliArgs
from axolotl.train import train
from axolotl.utils.config import normalize_config
from axolotl.utils.dict import DictDefault
LOG = logging.getLogger("axolotl.tests.e2e")
os.environ["WANDB_DISABLED"] = "true"
class TestLoraLlama(unittest.TestCase):
"""
Test case for Llama models using LoRA
"""
def test_lora(self):
# pylint: disable=duplicate-code
output_dir = tempfile.mkdtemp()
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"base_model_config": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
"lora_r": 32,
"lora_alpha": 64,
"lora_dropout": 0.05,
"lora_target_linear": True,
"val_set_size": 0.1,
"special_tokens": {
"unk_token": "<unk>",
"bos_token": "<s>",
"eos_token": "</s>",
},
"datasets": [
{
"path": "mhenrichsen/alpaca_2k_test",
"type": "alpaca",
},
],
"num_epochs": 2,
"micro_batch_size": 8,
"gradient_accumulation_steps": 1,
"output_dir": output_dir,
"learning_rate": 0.00001,
"optimizer": "adamw_torch",
"lr_scheduler": "cosine",
}
)
normalize_config(cfg)
cli_args = TrainerCliArgs()
dataset_meta = load_datasets(cfg=cfg, cli_args=cli_args)
train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta)
assert (Path(output_dir) / "adapter_model.bin").exists()
def test_lora_packing(self):
# pylint: disable=duplicate-code
output_dir = tempfile.mkdtemp()
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"base_model_config": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"sequence_len": 1024,
"sample_packing": True,
"flash_attention": True,
"load_in_8bit": True,
"adapter": "lora",
"lora_r": 32,
"lora_alpha": 64,
"lora_dropout": 0.05,
"lora_target_linear": True,
"val_set_size": 0.1,
"special_tokens": {
"unk_token": "<unk>",
"bos_token": "<s>",
"eos_token": "</s>",
},
"datasets": [
{
"path": "mhenrichsen/alpaca_2k_test",
"type": "alpaca",
},
],
"num_epochs": 2,
"micro_batch_size": 8,
"gradient_accumulation_steps": 1,
"output_dir": output_dir,
"learning_rate": 0.00001,
"optimizer": "adamw_torch",
"lr_scheduler": "cosine",
}
)
normalize_config(cfg)
cli_args = TrainerCliArgs()
dataset_meta = load_datasets(cfg=cfg, cli_args=cli_args)
train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta)
assert (Path(output_dir) / "adapter_model.bin").exists()
def test_lora_gptq(self):
# pylint: disable=duplicate-code
output_dir = tempfile.mkdtemp()
cfg = DictDefault(
{
"base_model": "TheBlokeAI/jackfram_llama-68m-GPTQ",
"base_model_config": "TheBlokeAI/jackfram_llama-68m-GPTQ",
"model_type": "AutoModelForCausalLM",
"tokenizer_type": "LlamaTokenizer",
"sequence_len": 1024,
"sample_packing": True,
"flash_attention": True,
"load_in_8bit": True,
"adapter": "lora",
"gptq": True,
"gptq_disable_exllama": True,
"lora_r": 32,
"lora_alpha": 64,
"lora_dropout": 0.05,
"lora_target_linear": True,
"val_set_size": 0.1,
"special_tokens": {
"unk_token": "<unk>",
"bos_token": "<s>",
"eos_token": "</s>",
},
"datasets": [
{
"path": "mhenrichsen/alpaca_2k_test",
"type": "alpaca",
},
],
"num_epochs": 2,
"save_steps": 0.5,
"micro_batch_size": 8,
"gradient_accumulation_steps": 1,
"output_dir": output_dir,
"learning_rate": 0.00001,
"optimizer": "adamw_torch",
"lr_scheduler": "cosine",
}
)
normalize_config(cfg)
cli_args = TrainerCliArgs()
dataset_meta = load_datasets(cfg=cfg, cli_args=cli_args)
train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta)
assert (Path(output_dir) / "adapter_model.bin").exists()
|