text string |
|---|
FINAL_ANSWER=RESULT(var=ANSWER1) |
torch.Size([7, 3, 448, 448]) |
dynamic ViT batch size: 13, images per sample: 13.0, dynamic token length: 3398 |
question: ['How many humans are in the image?'], responses:['2'] |
[('2', 0.12961991198727602), ('3', 0.12561270547489775), ('4', 0.12556127085987287), ('1', 0.1254920833223361), ('5', 0.12407835939022728), ('8', 0.124024076973589), ('7', 0.12288810153923228), ('29', 0.12272349045256851)] |
[['2', '3', '4', '1', '5', '8', '7', '29']] |
dynamic ViT batch size: 13, images per sample: 13.0, dynamic token length: 3398 |
torch.Size([7, 3, 448, 448]) knan debug pixel values shape |
tensor([8.8733e-01, 6.0383e-02, 8.6991e-03, 3.8982e-02, 3.2001e-03, 6.4994e-04, |
7.1398e-04, 3.9062e-05], device='cuda:0', grad_fn=<SoftmaxBackward0>) |
2 ************* |
['2', '3', '4', '1', '5', '8', '7', '29'] tensor([8.8733e-01, 6.0383e-02, 8.6991e-03, 3.8982e-02, 3.2001e-03, 6.4994e-04, |
7.1398e-04, 3.9062e-05], device='cuda:0', grad_fn=<SelectBackward0>) |
最后的概率分布为: {True: tensor(0.8873, device='cuda:0', grad_fn=<DivBackward0>), False: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>), 'Execute Error': tensor(-1.1921e-07, device='cuda:0', grad_fn=<DivBackward0>)} |
Encountered ExecuteError: CUDA out of memory. Tried to allocate 1020.00 MiB. GPU 2 has a total capacty of 44.34 GiB of which 946.94 MiB is free. Including non-PyTorch memory, this process has 43.40 GiB memory in use. Of the allocated memory 41.44 GiB is allocated by PyTorch, and 1.33 GiB is reserved by PyTorch but unal... |
Encountered TypeError: unsupported operand type(s) for +: 'NoneType' and 'str' |
最后的概率分布为: {True: 1e-09, False: 1e-09, 'Execute Error': 0.999999998} |
ANSWER0=VQA(image=RIGHT,question='How many pillows are in the image?') |
ANSWER1=EVAL(expr='{ANSWER0} == 2') |
FINAL_ANSWER=RESULT(var=ANSWER1) |
torch.Size([1, 3, 448, 448]) |
question: ['How many pillows are in the image?'], responses:['1'] |
[('1', 0.12829009354978346), ('3', 0.12529928082343206), ('4', 0.12464806219229535), ('8', 0.12460015878893425), ('6', 0.12451220062887247), ('12', 0.124338487048427), ('2', 0.12420459433498025), ('47', 0.12410712263327517)] |
[['1', '3', '4', '8', '6', '12', '2', '47']] |
torch.Size([1, 3, 448, 448]) knan debug pixel values shape |
tensor([9.8807e-01, 2.6068e-03, 1.4850e-03, 4.6540e-04, 7.9041e-04, 6.1294e-04, |
5.8753e-03, 9.2662e-05], device='cuda:2', grad_fn=<SoftmaxBackward0>) |
1 ************* |
['1', '3', '4', '8', '6', '12', '2', '47'] tensor([9.8807e-01, 2.6068e-03, 1.4850e-03, 4.6540e-04, 7.9041e-04, 6.1294e-04, |
5.8753e-03, 9.2662e-05], device='cuda:2', grad_fn=<SelectBackward0>) |
最后的概率分布为: {True: tensor(0.0059, device='cuda:2', grad_fn=<DivBackward0>), False: tensor(0.9941, device='cuda:2', grad_fn=<DivBackward0>), 'Execute Error': tensor(0., device='cuda:2', grad_fn=<DivBackward0>)} |
tensor([8.7346e-01, 6.3246e-02, 1.5991e-02, 4.0802e-02, 4.3045e-03, 9.9091e-04, |
1.1232e-03, 7.8810e-05], device='cuda:3', grad_fn=<SoftmaxBackward0>) |
2 ************* |
['2', '3', '4', '1', '5', '8', '7', '29'] tensor([8.7346e-01, 6.3246e-02, 1.5991e-02, 4.0802e-02, 4.3045e-03, 9.9091e-04, |
1.1232e-03, 7.8810e-05], device='cuda:3', grad_fn=<SelectBackward0>) |
最后的概率分布为: {True: tensor(0.8735, device='cuda:3', grad_fn=<DivBackward0>), False: tensor(0.1265, device='cuda:3', grad_fn=<DivBackward0>), 'Execute Error': tensor(5.9605e-08, device='cuda:3', grad_fn=<DivBackward0>)} |
petrel_client is not installed. If you read data locally instead of from ceph, ignore it. |
Replace train sampler!! |
petrel_client is not installed. Using PIL to load images. |
petrel_client is not installed. If you read data locally instead of from ceph, ignore it. |
Replace train sampler!! |
petrel_client is not installed. Using PIL to load images. |
petrel_client is not installed. If you read data locally instead of from ceph, ignore it. |
Replace train sampler!! |
petrel_client is not installed. Using PIL to load images. |
petrel_client is not installed. If you read data locally instead of from ceph, ignore it. |
Replace train sampler!! |
petrel_client is not installed. Using PIL to load images. |
Traceback (most recent call last): |
File "/mnt/SSD1_4TB/yunjie/Internvl_NLVR/InternVL/internvl_chat/internvl/train/internvl_chat_finetune.py", line 898, in <module> |
main() |
File "/mnt/SSD1_4TB/yunjie/Internvl_NLVR/InternVL/internvl_chat/internvl/train/internvl_chat_finetune.py", line 882, in main |
train_result = trainer.train(resume_from_checkpoint=checkpoint) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/transformers/trainer.py", line 1539, in train |
return inner_training_loop( |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/transformers/trainer.py", line 1869, in _inner_training_loop |
tr_loss_step = self.training_step(model, inputs) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/transformers/trainer.py", line 2783, in training_step |
self.accelerator.backward(loss) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/accelerate/accelerator.py", line 2188, in backward |
self.deepspeed_engine_wrapped.backward(loss, **kwargs) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/accelerate/utils/deepspeed.py", line 166, in backward |
self.engine.backward(loss, **kwargs) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/deepspeed/utils/nvtx.py", line 15, in wrapped_fn |
ret_val = func(*args, **kwargs) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/deepspeed/runtime/engine.py", line 1895, in backward |
self.optimizer.backward(loss, retain_graph=retain_graph) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/deepspeed/runtime/zero/stage_1_and_2.py", line 1902, in backward |
self.loss_scaler.backward(loss.float(), retain_graph=retain_graph) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/deepspeed/runtime/fp16/loss_scaler.py", line 63, in backward |
scaled_loss.backward(retain_graph=retain_graph) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/_tensor.py", line 492, in backward |
torch.autograd.backward( |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/autograd/__init__.py", line 251, in backward |
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/autograd/function.py", line 288, in apply |
return user_fn(self, *args) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/utils/checkpoint.py", line 288, in backward |
torch.autograd.backward(outputs_with_grad, args_with_grad) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/autograd/__init__.py", line 251, in backward |
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass |
torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 2.00 MiB. GPU 3 has a total capacty of 44.34 GiB of which 2.94 MiB is free. Including non-PyTorch memory, this process has 44.32 GiB memory in use. Of the allocated memory 41.03 GiB is allocated by PyTorch, and 2.74 GiB is reserved by PyTorch but unallo... |
[2024-10-22 17:32:36,618] torch.distributed.elastic.multiprocessing.api: [WARNING] Sending process 2084211 closing signal SIGTERM |
[2024-10-22 17:32:36,618] torch.distributed.elastic.multiprocessing.api: [WARNING] Sending process 2084212 closing signal SIGTERM |
[2024-10-22 17:32:36,619] torch.distributed.elastic.multiprocessing.api: [WARNING] Sending process 2084213 closing signal SIGTERM |
[2024-10-22 17:32:38,236] torch.distributed.elastic.multiprocessing.api: [ERROR] failed (exitcode: 1) local_rank: 3 (pid: 2084214) of binary: /home/yunjie/anaconda3/envs/internvl/bin/python |
Traceback (most recent call last): |
File "/home/yunjie/anaconda3/envs/internvl/bin/torchrun", line 8, in <module> |
sys.exit(main()) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper |
return f(*args, **kwargs) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/distributed/run.py", line 806, in main |
run(args) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/distributed/run.py", line 797, in run |
elastic_launch( |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 134, in __call__ |
return launch_agent(self._config, self._entrypoint, list(args)) |
File "/home/yunjie/anaconda3/envs/internvl/lib/python3.10/site-packages/torch/distributed/launcher/api.py", line 264, in launch_agent |
raise ChildFailedError( |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.