00cda8cc70
* data loading refactor (wip) * updates * progress * pytest * pytest fix * lint * zero_first -> filelock, more simplifications * small simplification * import change * nit * lint * simplify dedup * couldnt resist * review comments WIP * continued wip * minor changes * fix; remove contrived test * further refactor * set default seed in pydantic config * lint * continued simplication * lint * renaming and nits * filelock tests * fix * fix * lint * remove nullable arg * remove unnecessary code * moving dataset save fn to shared module * remove debug print * matching var naming * fn name change * coderabbit comments * naming nit * fix test
188 lines
6.6 KiB
Python
188 lines
6.6 KiB
Python
"""
|
|
E2E tests for llama
|
|
"""
|
|
|
|
from axolotl.common.datasets import load_datasets
|
|
from axolotl.train import train
|
|
from axolotl.utils.config import normalize_config, validate_config
|
|
from axolotl.utils.dict import DictDefault
|
|
|
|
from tests.e2e.utils import check_model_output_exists
|
|
|
|
|
|
class TestLlama:
|
|
"""
|
|
Test case for Llama models
|
|
"""
|
|
|
|
def test_fft_trust_remote_code(self, temp_dir):
|
|
# pylint: disable=duplicate-code
|
|
cfg = DictDefault(
|
|
{
|
|
"base_model": "HuggingFaceTB/SmolLM2-135M",
|
|
"tokenizer_type": "AutoTokenizer",
|
|
"trust_remote_code": True,
|
|
"sequence_len": 512,
|
|
"val_set_size": 0.02,
|
|
"special_tokens": {
|
|
"pad_token": "<|endoftext|>",
|
|
},
|
|
"datasets": [
|
|
{
|
|
"path": "mhenrichsen/alpaca_2k_test",
|
|
"type": "alpaca",
|
|
},
|
|
],
|
|
"num_epochs": 1,
|
|
"max_steps": 5,
|
|
"micro_batch_size": 2,
|
|
"gradient_accumulation_steps": 1,
|
|
"output_dir": temp_dir,
|
|
"learning_rate": 0.00001,
|
|
"optimizer": "adamw_bnb_8bit",
|
|
"lr_scheduler": "cosine",
|
|
"flash_attention": True,
|
|
"sample_packing": True,
|
|
"bf16": True,
|
|
"save_safetensors": True,
|
|
}
|
|
)
|
|
|
|
cfg = validate_config(cfg)
|
|
normalize_config(cfg)
|
|
dataset_meta = load_datasets(cfg=cfg)
|
|
|
|
train(cfg=cfg, dataset_meta=dataset_meta)
|
|
check_model_output_exists(temp_dir, cfg)
|
|
|
|
def test_fix_untrained_tokens(self, temp_dir):
|
|
# pylint: disable=duplicate-code
|
|
cfg = DictDefault(
|
|
{
|
|
"base_model": "HuggingFaceTB/SmolLM2-135M",
|
|
"fix_untrained_tokens": True,
|
|
"sequence_len": 512,
|
|
"val_set_size": 0.0,
|
|
"special_tokens": {
|
|
"pad_token": "<|endoftext|>",
|
|
"bos_token": "<|custom_im_start|>",
|
|
"eos_token": "<|custom_im_end|>",
|
|
},
|
|
"datasets": [
|
|
{
|
|
"chat_template": "jinja",
|
|
"chat_template_jinja": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in messages %}{{'<|custom_im_start|>' + message['role'] + '\n' + message['content'] + '<|custom_im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|custom_im_start|>assistant\n' }}{% endif %}",
|
|
"path": "mlabonne/FineTome-100k",
|
|
"type": "chat_template",
|
|
"split": "train[:10%]",
|
|
"field_messages": "conversations",
|
|
"message_field_role": "from",
|
|
"message_field_content": "value",
|
|
},
|
|
],
|
|
"num_epochs": 1,
|
|
"max_steps": 5,
|
|
"micro_batch_size": 1,
|
|
"gradient_accumulation_steps": 1,
|
|
"output_dir": temp_dir,
|
|
"learning_rate": 0.00001,
|
|
"optimizer": "adamw_8bit",
|
|
"lr_scheduler": "cosine",
|
|
"flash_attention": True,
|
|
"sample_packing": True,
|
|
"bf16": True,
|
|
"save_safetensors": True,
|
|
}
|
|
)
|
|
|
|
cfg = validate_config(cfg)
|
|
normalize_config(cfg)
|
|
dataset_meta = load_datasets(cfg=cfg)
|
|
|
|
train(cfg=cfg, dataset_meta=dataset_meta)
|
|
check_model_output_exists(temp_dir, cfg)
|
|
|
|
def test_fix_untrained_tokens_already_trained(self, temp_dir):
|
|
# pylint: disable=duplicate-code
|
|
cfg = DictDefault(
|
|
{
|
|
"base_model": "HuggingFaceTB/SmolLM2-135M",
|
|
"fix_untrained_tokens": True,
|
|
"sequence_len": 512,
|
|
"val_set_size": 0.0,
|
|
"special_tokens": {
|
|
"pad_token": "<|endoftext|>",
|
|
},
|
|
"chat_template": "chatml",
|
|
"datasets": [
|
|
{
|
|
"path": "mlabonne/FineTome-100k",
|
|
"type": "chat_template",
|
|
"split": "train[:10%]",
|
|
"field_messages": "conversations",
|
|
"message_field_role": "from",
|
|
"message_field_content": "value",
|
|
},
|
|
],
|
|
"num_epochs": 1,
|
|
"max_steps": 5,
|
|
"micro_batch_size": 1,
|
|
"gradient_accumulation_steps": 1,
|
|
"output_dir": temp_dir,
|
|
"learning_rate": 0.00001,
|
|
"optimizer": "adamw_8bit",
|
|
"lr_scheduler": "cosine",
|
|
"flash_attention": True,
|
|
"sample_packing": True,
|
|
"bf16": True,
|
|
"save_safetensors": True,
|
|
}
|
|
)
|
|
|
|
cfg = validate_config(cfg)
|
|
normalize_config(cfg)
|
|
dataset_meta = load_datasets(cfg=cfg)
|
|
|
|
train(cfg=cfg, dataset_meta=dataset_meta)
|
|
check_model_output_exists(temp_dir, cfg)
|
|
|
|
def test_batch_flattening(self, temp_dir):
|
|
# pylint: disable=duplicate-code
|
|
cfg = DictDefault(
|
|
{
|
|
"base_model": "HuggingFaceTB/SmolLM2-135M",
|
|
"trust_remote_code": True,
|
|
"sequence_len": 512,
|
|
"val_set_size": 0.01,
|
|
"special_tokens": {
|
|
"pad_token": "<|endoftext|>",
|
|
},
|
|
"datasets": [
|
|
{
|
|
"path": "mhenrichsen/alpaca_2k_test",
|
|
"type": "alpaca",
|
|
},
|
|
],
|
|
"num_epochs": 1,
|
|
"max_steps": 5,
|
|
"micro_batch_size": 4,
|
|
"gradient_accumulation_steps": 1,
|
|
"output_dir": temp_dir,
|
|
"learning_rate": 0.00001,
|
|
"optimizer": "adamw_8bit",
|
|
"lr_scheduler": "cosine",
|
|
"flash_attention": True,
|
|
"sample_packing": False,
|
|
"batch_flattening": True,
|
|
"bf16": True,
|
|
"save_safetensors": True,
|
|
}
|
|
)
|
|
|
|
cfg = validate_config(cfg)
|
|
normalize_config(cfg)
|
|
dataset_meta = load_datasets(cfg=cfg)
|
|
|
|
train(cfg=cfg, dataset_meta=dataset_meta)
|
|
check_model_output_exists(temp_dir, cfg)
|