File size: 7,349 Bytes
7523d1f 0001862 7523d1f 0001862 7523d1f 98c25e1 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 |
"""
E2E tests for lora llama
"""
import logging
import os
import unittest
from pathlib import Path
import pytest
from axolotl.cli import load_rl_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
from .utils import with_temp_dir
LOG = logging.getLogger("axolotl.tests.e2e")
os.environ["WANDB_DISABLED"] = "true"
@pytest.mark.skip(reason="doesn't seem to work on modal")
class TestDPOLlamaLora(unittest.TestCase):
"""
Test case for DPO Llama models using LoRA
"""
@with_temp_dir
def test_dpo_lora(self, temp_dir):
# pylint: disable=duplicate-code
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
"lora_r": 64,
"lora_alpha": 32,
"lora_dropout": 0.1,
"lora_target_linear": True,
"special_tokens": {},
"rl": "dpo",
"datasets": [
{
"path": "Intel/orca_dpo_pairs",
"type": "chatml.intel",
"split": "train",
},
],
"num_epochs": 1,
"micro_batch_size": 4,
"gradient_accumulation_steps": 1,
"output_dir": temp_dir,
"learning_rate": 0.00001,
"optimizer": "paged_adamw_8bit",
"lr_scheduler": "cosine",
"max_steps": 20,
"save_steps": 10,
"warmup_steps": 5,
"gradient_checkpointing": True,
"gradient_checkpointing_kwargs": {"use_reentrant": True},
}
)
normalize_config(cfg)
cli_args = TrainerCliArgs()
dataset_meta = load_rl_datasets(cfg=cfg, cli_args=cli_args)
train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta)
assert (Path(temp_dir) / "checkpoint-20/adapter_model.safetensors").exists()
@with_temp_dir
def test_kto_pair_lora(self, temp_dir):
# pylint: disable=duplicate-code
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
"lora_r": 64,
"lora_alpha": 32,
"lora_dropout": 0.1,
"lora_target_linear": True,
"special_tokens": {},
"rl": "kto_pair",
"datasets": [
{
"path": "Intel/orca_dpo_pairs",
"type": "chatml.intel",
"split": "train",
},
],
"num_epochs": 1,
"micro_batch_size": 4,
"gradient_accumulation_steps": 1,
"output_dir": temp_dir,
"learning_rate": 0.00001,
"optimizer": "paged_adamw_8bit",
"lr_scheduler": "cosine",
"max_steps": 20,
"save_steps": 10,
"warmup_steps": 5,
"gradient_checkpointing": True,
"gradient_checkpointing_kwargs": {"use_reentrant": True},
}
)
normalize_config(cfg)
cli_args = TrainerCliArgs()
dataset_meta = load_rl_datasets(cfg=cfg, cli_args=cli_args)
train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta)
assert (Path(temp_dir) / "checkpoint-20/adapter_model.safetensors").exists()
@with_temp_dir
def test_ipo_lora(self, temp_dir):
# pylint: disable=duplicate-code
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
"lora_r": 64,
"lora_alpha": 32,
"lora_dropout": 0.1,
"lora_target_linear": True,
"special_tokens": {},
"rl": "ipo",
"datasets": [
{
"path": "Intel/orca_dpo_pairs",
"type": "chatml.intel",
"split": "train",
},
],
"num_epochs": 1,
"micro_batch_size": 4,
"gradient_accumulation_steps": 1,
"output_dir": temp_dir,
"learning_rate": 0.00001,
"optimizer": "paged_adamw_8bit",
"lr_scheduler": "cosine",
"max_steps": 20,
"save_steps": 10,
"warmup_steps": 5,
"gradient_checkpointing": True,
"gradient_checkpointing_kwargs": {"use_reentrant": True},
}
)
normalize_config(cfg)
cli_args = TrainerCliArgs()
dataset_meta = load_rl_datasets(cfg=cfg, cli_args=cli_args)
train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta)
assert (Path(temp_dir) / "checkpoint-20/adapter_model.safetensors").exists()
@with_temp_dir
def test_orpo_lora(self, temp_dir):
# pylint: disable=duplicate-code
cfg = DictDefault(
{
"base_model": "JackFram/llama-68m",
"tokenizer_type": "LlamaTokenizer",
"sequence_len": 1024,
"load_in_8bit": True,
"adapter": "lora",
"lora_r": 64,
"lora_alpha": 32,
"lora_dropout": 0.1,
"lora_target_linear": True,
"special_tokens": {},
"rl": "orpo",
"orpo_alpha": 0.1,
"remove_unused_columns": False,
"chat_template": "chatml",
"datasets": [
{
"path": "argilla/ultrafeedback-binarized-preferences-cleaned",
"type": "chat_template.argilla",
"split": "train",
},
],
"num_epochs": 1,
"micro_batch_size": 4,
"gradient_accumulation_steps": 1,
"output_dir": temp_dir,
"learning_rate": 0.00001,
"optimizer": "paged_adamw_8bit",
"lr_scheduler": "cosine",
"max_steps": 20,
"save_steps": 10,
"warmup_steps": 5,
"gradient_checkpointing": True,
"gradient_checkpointing_kwargs": {"use_reentrant": True},
}
)
normalize_config(cfg)
cli_args = TrainerCliArgs()
dataset_meta = load_rl_datasets(cfg=cfg, cli_args=cli_args)
train(cfg=cfg, cli_args=cli_args, dataset_meta=dataset_meta)
assert (Path(temp_dir) / "checkpoint-20/adapter_model.safetensors").exists()
|