[BUG] AttributeError: module 'torch.library' has no attribute 'register_fake'

[azmeer36]

Prerequisites

• I have read the documentation.
• I have checked other issues for similar problems.

Backend

Colab

Interface Used

CLI

CLI Command

No response

UI Screenshots & Parameters

No response

Error Logs

INFO | 2024-08-20 22:32:17 | autotrain.cli.autotrain:main:60 - Using AutoTrain configuration: conf.yaml
INFO | 2024-08-20 22:32:17 | autotrain.parser:post_init:147 - Running task: lm_training
INFO | 2024-08-20 22:32:17 | autotrain.parser:post_init:148 - Using backend: local
INFO | 2024-08-20 22:32:17 | autotrain.parser:run:211 - {'model': 'abhishek/llama-2-7b-hf-small-shards', 'project_name': 'my-autotrain-llm', 'data_path': '/content', 'train_split': 'train', 'valid_split': None, 'add_eos_token': True, 'block_size': 1024, 'model_max_length': 2048, 'padding': 'right', 'trainer': 'sft', 'use_flash_attention_2': False, 'log': 'tensorboard', 'disable_gradient_checkpointing': False, 'logging_steps': -1, 'eval_strategy': 'epoch', 'save_total_limit': 1, 'auto_find_batch_size': False, 'mixed_precision': 'fp16', 'lr': 0.0002, 'epochs': 1, 'batch_size': 1, 'warmup_ratio': 0.1, 'gradient_accumulation': 4, 'optimizer': 'adamw_torch', 'scheduler': 'linear', 'weight_decay': 0.01, 'max_grad_norm': 1.0, 'seed': 42, 'chat_template': None, 'quantization': 'int4', 'target_modules': 'all-linear', 'merge_adapter': False, 'peft': True, 'lora_r': 16, 'lora_alpha': 32, 'lora_dropout': 0.05, 'model_ref': None, 'dpo_beta': 0.1, 'max_prompt_length': 128, 'max_completion_length': None, 'prompt_text_column': None, 'text_column': 'text', 'rejected_text_column': None, 'push_to_hub': True, 'username': 'AzmeerFaisal', 'token': '', 'unsloth': False}
Saving the dataset (1/1 shards): 100% 9846/9846 [00:00<00:00, 228659.88 examples/s]
Saving the dataset (1/1 shards): 100% 9846/9846 [00:00<00:00, 259192.35 examples/s]
INFO | 2024-08-20 22:32:18 | autotrain.backends.local:create:8 - Starting local training...
WARNING | 2024-08-20 22:32:18 | autotrain.commands:launch_command:47 - No GPU found. Forcing training on CPU. This will be super slow!
INFO | 2024-08-20 22:32:18 | autotrain.commands:launch_command:489 - ['accelerate', 'launch', '--cpu', '-m', 'autotrain.trainers.clm', '--training_config', 'my-autotrain-llm/training_params.json']
INFO | 2024-08-20 22:32:18 | autotrain.commands:launch_command:490 - {'model': 'abhishek/llama-2-7b-hf-small-shards', 'project_name': 'my-autotrain-llm', 'data_path': 'my-autotrain-llm/autotrain-data', 'train_split': 'train', 'valid_split': None, 'add_eos_token': True, 'block_size': 1024, 'model_max_length': 2048, 'padding': 'right', 'trainer': 'sft', 'use_flash_attention_2': False, 'log': 'tensorboard', 'disable_gradient_checkpointing': False, 'logging_steps': -1, 'eval_strategy': 'epoch', 'save_total_limit': 1, 'auto_find_batch_size': False, 'mixed_precision': 'fp16', 'lr': 0.0002, 'epochs': 1, 'batch_size': 1, 'warmup_ratio': 0.1, 'gradient_accumulation': 4, 'optimizer': 'adamw_torch', 'scheduler': 'linear', 'weight_decay': 0.01, 'max_grad_norm': 1.0, 'seed': 42, 'chat_template': None, 'quantization': 'int4', 'target_modules': 'all-linear', 'merge_adapter': False, 'peft': True, 'lora_r': 16, 'lora_alpha': 32, 'lora_dropout': 0.05, 'model_ref': None, 'dpo_beta': 0.1, 'max_prompt_length': 128, 'max_completion_length': None, 'prompt_text_column': 'autotrain_prompt', 'text_column': 'autotrain_text', 'rejected_text_column': 'autotrain_rejected_text', 'push_to_hub': True, 'username': 'AzmeerFaisal', 'token': '', 'unsloth': False}
Traceback (most recent call last):
File "/usr/local/bin/accelerate", line 5, in
from accelerate.commands.accelerate_cli import main
File "/usr/local/lib/python3.10/dist-packages/accelerate/commands/accelerate_cli.py", line 19, in
from accelerate.commands.estimate import estimate_command_parser
File "/usr/local/lib/python3.10/dist-packages/accelerate/commands/estimate.py", line 34, in
import timm
File "/usr/local/lib/python3.10/dist-packages/timm/init.py", line 2, in
from .layers import is_scriptable, is_exportable, set_scriptable, set_exportable
File "/usr/local/lib/python3.10/dist-packages/timm/layers/init.py", line 8, in
from .classifier import ClassifierHead, create_classifier, NormMlpClassifierHead
File "/usr/local/lib/python3.10/dist-packages/timm/layers/classifier.py", line 15, in
from .create_norm import get_norm_layer
File "/usr/local/lib/python3.10/dist-packages/timm/layers/create_norm.py", line 14, in
from torchvision.ops.misc import FrozenBatchNorm2d
File "/usr/local/lib/python3.10/dist-packages/torchvision/init.py", line 10, in
from torchvision import _meta_registrations, datasets, io, models, ops, transforms, utils # usort:skip
File "/usr/local/lib/python3.10/dist-packages/torchvision/_meta_registrations.py", line 163, in
@torch.library.register_fake("torchvision::nms")
AttributeError: module 'torch.library' has no attribute 'register_fake'
INFO | 2024-08-20 22:32:22 | autotrain.parser:run:216 - Job ID: 20950

Additional Information

No response

[abhishekkrthakur]

whats your torch version?

[azmeer36]

whats your torch version?

Torch version: 2.2.0+cu121

This is my first cell code:
import os
!pip install -U autotrain-advanced > install_logs.txt 2>&1
!pip install -U torch torchvision
!autotrain setup --update-torch
!autotrain setup --colab > setup_logs.txt
from autotrain import version
print(f'AutoTrain version: {version}')
from torch import version
print(f'Torch version: {version}')

[abhishekkrthakur]

please use pytorch==2.3.0

[azmeer36]

I have tried running this:

import os
!pip install -U autotrain-advanced > install_logs.txt 2>&1
!pip install torch==2.3.0 torchvision
!autotrain setup --colab > setup_logs.txt
from autotrain import version
print(f'AutoTrain version: {version}')
from torch import version
print(f'Torch version: {version}')

It still says:
Torch version: 2.2.0+cu121

[ngwgsang]

@azmeer36
try installing torch==2.4.0 , it worked in my case.

import os
!pip install -U autotrain-advanced > install_logs.txt 2>&1
!autotrain setup --colab > setup_logs.txt
!pip install torch==2.4.0
from autotrain import __version__
import torch

[github-actions]

This issue is stale because it has been open for 30 days with no activity.