b6fc46ada8
* add grpo scale_rewards config for trl#3135 * options to connect to vllm server directly w grpo trl#3094 * temperature support trl#3029 * sampling/generation kwargs for grpo trl#2989 * make vllm_enable_prefix_caching a config param trl#2900 * grpo multi-step optimizeations trl#2899 * remove overrides for grpo trainer * bump trl to 0.16.0 * add cli to start vllm-serve via trl * call the python module directly * update to use vllm with 2.6.0 too now and call trl vllm serve from module * vllm 0.8.1 * use python3 * use sys.executable * remove context and wait for start * fixes to make it actually work * fixes so the grpo tests pass with new vllm paradigm * explicit host/port and check in start vllm * make sure that vllm doesn't hang by setting quiet so outouts go to dev null * also bump bnb to latest release * add option for wait from cli and nccl debugging for ci * grpo + vllm test on separate devices for now * make sure grpo + vllm tests runs single worker since pynccl comms would conflict * fix cli * remove wait and add caching for argilla dataset * refactoring configs * chore: lint * add vllm config * fixup vllm grpo args * fix one more incorrect schema/config path * fix another vlllm reference and increase timeout * make the tests run a bit faster * change mbsz back so it is correct for grpo * another change mbsz back so it is correct for grpo * fixing cli args * nits * adding docs * docs * include tensor parallel size for vllm in pydantic schema * moving start_vllm, more docs * limit output len for grpo vllm * vllm enable_prefix_caching isn't a bool cli arg * fix env ordering in tests and also use pid check when looking for vllm --------- Co-authored-by: Salman Mohammadi <salman.mohammadi@outlook.com>
98 lines
3.2 KiB
Python
98 lines
3.2 KiB
Python
"""
|
|
E2E tests for multigpu qwen2
|
|
"""
|
|
|
|
import logging
|
|
import os
|
|
from pathlib import Path
|
|
|
|
import pytest
|
|
import yaml
|
|
from accelerate.test_utils import execute_subprocess_async
|
|
from transformers.testing_utils import get_torch_dist_unique_port
|
|
|
|
from axolotl.utils.dict import DictDefault
|
|
|
|
LOG = logging.getLogger("axolotl.tests.e2e.multigpu")
|
|
os.environ["WANDB_DISABLED"] = "true"
|
|
|
|
|
|
class TestMultiGPUQwen2:
|
|
"""
|
|
Test case for Llama models using LoRA
|
|
"""
|
|
|
|
@pytest.mark.parametrize("base_model", ["Qwen/Qwen2-0.5B", "Qwen/Qwen2.5-0.5B"])
|
|
def test_qlora_fsdp_dpo(self, base_model, temp_dir):
|
|
# pylint: disable=duplicate-code
|
|
cfg = DictDefault(
|
|
{
|
|
"base_model": base_model,
|
|
"load_in_4bit": True,
|
|
"rl": "dpo",
|
|
"chat_template": "chatml",
|
|
"sequence_len": 2048,
|
|
"adapter": "qlora",
|
|
"lora_r": 8,
|
|
"lora_alpha": 16,
|
|
"lora_dropout": 0.05,
|
|
"lora_target_linear": True,
|
|
"val_set_size": 0.05,
|
|
"datasets": [
|
|
{
|
|
"path": "Intel/orca_dpo_pairs",
|
|
"split": "train",
|
|
"type": "chatml.intel",
|
|
},
|
|
],
|
|
"num_epochs": 1,
|
|
"max_steps": 2,
|
|
"warmup_steps": 20,
|
|
"micro_batch_size": 2,
|
|
"gradient_accumulation_steps": 2,
|
|
"output_dir": temp_dir,
|
|
"learning_rate": 0.00001,
|
|
"optimizer": "adamw_torch_fused",
|
|
"lr_scheduler": "cosine",
|
|
"flash_attention": True,
|
|
"bf16": "auto",
|
|
"tf32": True,
|
|
"gradient_checkpointing": True,
|
|
"gradient_checkpointing_kwargs": {
|
|
"use_reentrant": False,
|
|
},
|
|
"fsdp": [
|
|
"full_shard",
|
|
"auto_wrap",
|
|
],
|
|
"fsdp_config": {
|
|
"fsdp_limit_all_gathers": True,
|
|
"fsdp_offload_params": False,
|
|
"fsdp_sync_module_states": True,
|
|
"fsdp_use_orig_params": False,
|
|
"fsdp_cpu_ram_efficient_loading": False,
|
|
"fsdp_transformer_layer_cls_to_wrap": "Qwen2DecoderLayer",
|
|
"fsdp_state_dict_type": "FULL_STATE_DICT",
|
|
"fsdp_auto_wrap_policy": "TRANSFORMER_BASED_WRAP",
|
|
"fsdp_sharding_strategy": "FULL_SHARD",
|
|
},
|
|
}
|
|
)
|
|
|
|
# write cfg to yaml file
|
|
Path(temp_dir).mkdir(parents=True, exist_ok=True)
|
|
with open(Path(temp_dir) / "config.yaml", "w", encoding="utf-8") as fout:
|
|
fout.write(yaml.dump(cfg.to_dict(), Dumper=yaml.Dumper))
|
|
|
|
execute_subprocess_async(
|
|
[
|
|
"axolotl",
|
|
"train",
|
|
str(Path(temp_dir) / "config.yaml"),
|
|
"--num-processes",
|
|
"2",
|
|
"--main-process-port",
|
|
f"{get_torch_dist_unique_port()}",
|
|
]
|
|
)
|