phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-08-03 17:29:01 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

pre-commit

Browse source
This commit is contained in:
Botao Chen 2025-01-02 20:23:55 -08:00
parent dadb257f03
commit 562d5228de
2 changed files with 21 additions and 20 deletions
Download patch file Download diff file Expand all files Collapse all files

5
llama_stack/providers/inline/post_training/torchtune/common/utils.py
View file

@ -16,8 +16,6 @@ from typing import Any, Callable, Dict, List
import torch import torch
from llama_models.datatypes import Model from llama_models.datatypes import Model
from llama_models.sku_list import resolve_model from llama_models.sku_list import resolve_model
from llama_stack.apis.common.type_system import ParamType, StringType
from llama_stack.apis.datasets import Datasets
from pydantic import BaseModel from pydantic import BaseModel
@ -25,6 +23,9 @@ from torchtune.models.llama3 import llama3_tokenizer, lora_llama3_8b
from torchtune.models.llama3._tokenizer import Llama3Tokenizer from torchtune.models.llama3._tokenizer import Llama3Tokenizer
from torchtune.models.llama3_2 import lora_llama3_2_3b from torchtune.models.llama3_2 import lora_llama3_2_3b
from llama_stack.apis.common.type_system import ParamType, StringType
from llama_stack.apis.datasets import Datasets
class ColumnName(Enum): class ColumnName(Enum):
instruction = "instruction" instruction = "instruction"

36
llama_stack/providers/inline/post_training/torchtune/recipes/lora_finetuning_single_device.py
View file

@ -14,6 +14,24 @@ from typing import Any, Dict, List, Optional, Tuple
import torch import torch
from llama_models.sku_list import resolve_model from llama_models.sku_list import resolve_model
from torch import nn
from torch.optim import Optimizer
from torch.utils.data import DataLoader, DistributedSampler
from torchtune import modules, training, utils as torchtune_utils
from torchtune.data import AlpacaToMessages, padded_collate_sft
from torchtune.modules.loss import CEWithChunkedOutputLoss
from torchtune.modules.peft import (
get_adapter_params,
get_adapter_state_dict,
get_lora_module_names,
get_merged_lora_ckpt,
set_trainable_params,
validate_missing_and_unexpected_for_lora,
)
from torchtune.training.lr_schedulers import get_cosine_schedule_with_warmup
from torchtune.training.metric_logging import DiskLogger
from tqdm import tqdm
from llama_stack.apis.common.training_types import PostTrainingMetric from llama_stack.apis.common.training_types import PostTrainingMetric
from llama_stack.apis.datasetio import DatasetIO from llama_stack.apis.datasetio import DatasetIO
@ -38,24 +56,6 @@ from llama_stack.providers.inline.post_training.torchtune.config import (
TorchtunePostTrainingConfig, TorchtunePostTrainingConfig,
) )
from llama_stack.providers.inline.post_training.torchtune.datasets.sft import SFTDataset from llama_stack.providers.inline.post_training.torchtune.datasets.sft import SFTDataset
from torch import nn
from torch.optim import Optimizer
from torch.utils.data import DataLoader, DistributedSampler
from torchtune import modules, training, utils as torchtune_utils
from torchtune.data import AlpacaToMessages, padded_collate_sft
from torchtune.modules.loss import CEWithChunkedOutputLoss
from torchtune.modules.peft import (
get_adapter_params,
get_adapter_state_dict,
get_lora_module_names,
get_merged_lora_ckpt,
set_trainable_params,
validate_missing_and_unexpected_for_lora,
)
from torchtune.training.lr_schedulers import get_cosine_schedule_with_warmup
from torchtune.training.metric_logging import DiskLogger
from tqdm import tqdm
log = logging.getLogger(__name__) log = logging.getLogger(__name__)