Comments (9)
@adarob could you please help me out on this issue
from t5x.
I think your seqio version is out of date. There was a recent change that added trim_output_features
(https://github.com/google/seqio/blob/90c76914ed13fcce53f00966b824e45fb266b973/seqio/dataset_providers.py#L1123).
Can you update the seqio and rerun?
from t5x.
@hwchung27 yeah on updating the seqio the error was cleared. however upon running the training script locally on 2 x A6000 GPU. intially jax takes too long to compile and only one gpu get completely full while the other remains empty. and they the program crashes.
from t5x.
@hwchung27 i am pretraining sentencepiece tokenizer because the mt5 model i plan to pretrain contains languages with language_scripts that the mt5 model doesnt cover.
However, ive noticed that all t5 model use/reserve 100 ids/vocabs. Hence when training the sentencepiece tokenizers should i also train them with +100 tokens ?
i mean given mt5 has vocab of 250000 --> so should i train for 250100 ?
eg: spm_train --input=alltext.txt --model_prefix=mt5_tokenizer_250100_unigram --character_coverage=0.9995 --model_type=unigram --vocab_size=250100
from t5x.
Nope you don't need to add extra ids. The seqio.SentencePieceVocabulary
does the "surgery" for you: https://github.com/google/seqio/blob/71e47acd6fd38d0cb1b89ce112df56f37e054f2c/seqio/vocabularies.py#L304-L321
from t5x.
Thanks a ton for replying.
If you don't mind, could you also please elaborate on why and what's the purpose of adding +100 ids in t5 ?
from t5x.
i am using my own tokenizer for the pretraining task:
however, when i add extra_ids=100
to the seqio.SentencePieceVocabulary()
then i get the error:
raise ValueError(f'Model and Task vocabularies do not match:\n'
ValueError: Model and Task vocabularies do not match:
note: i have changed the default tokenizer to my own tokenizer in the base.gin
file
however upon changing the extra_ids=0 this error resolved but i get this error:
ValueError: None values not supported.
the following is the entire stack track of the error:
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/util/traceback_utils.py", line 153, in error_handler
raise e.with_traceback(filtered_tb) from None
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/framework/tensor_util.py", line 441, in make_tensor_proto
raise ValueError("None values not supported.")
ValueError: in user code:
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/utils.py", line 779, in None *
lambda arg: fn(arg, *args, **kargs)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/preprocessors.py", line 116, in tokenize_impl *
v = vocab.encode_tf(v)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/vocabularies.py", line 114, in encode_tf *
return self._encode_tf(s)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/vocabularies.py", line 413, in _encode_tf *
return self.tf_tokenizer.tokenize(s)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow_text/python/ops/sentencepiece_tokenizer.py", line 133, in tokenize *
input_tensor = ragged_tensor.convert_to_tensor_or_ragged_tensor(input)
ValueError: None values not supported.
In call to configurable 'train' (<function train at 0x7f1245e43280>)
from t5x.
@hwchung27 Just an update on the same.
I use the default mt5 tokenizer gs://t5-data/vocabs/mc4.250000.100extra/sentencepiece.model
witrh extra_ids=0
and i stil get the same error.
The following is the entire stack track of the same:
Traceback (most recent call last):
File "/home/stephen/anaconda3/lib/python3.9/runpy.py", line 197, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/stephen/anaconda3/lib/python3.9/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/stephen/Desktop/t5x_final_test/t5x/t5x/train.py", line 748, in <module>
gin_utils.run(main)
File "/home/stephen/Desktop/t5x_final_test/t5x/t5x/gin_utils.py", line 107, in run
app.run(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/absl/app.py", line 308, in run
_run_main(main, args)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/absl/app.py", line 254, in _run_main
sys.exit(main(argv))
File "/home/stephen/Desktop/t5x_final_test/t5x/t5x/train.py", line 708, in main
_main(argv)
File "/home/stephen/Desktop/t5x_final_test/t5x/t5x/train.py", line 744, in _main
train_using_gin()
File "/home/stephen/anaconda3/lib/python3.9/site-packages/gin/config.py", line 1605, in gin_wrapper
utils.augment_exception_message_and_reraise(e, err_str)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/gin/utils.py", line 41, in augment_exception_message_and_reraise
raise proxy.with_traceback(exception.__traceback__) from None
File "/home/stephen/anaconda3/lib/python3.9/site-packages/gin/config.py", line 1582, in gin_wrapper
return fn(*new_args, **new_kwargs)
File "/home/stephen/Desktop/t5x_final_test/t5x/t5x/train.py", line 249, in train
train_ds = get_dataset_fn(train_dataset_cfg, ds_shard_id, num_ds_shards,
File "/home/stephen/Desktop/t5x_final_test/t5x/t5x/utils.py", line 1366, in get_dataset
return get_dataset_inner(cfg, shard_info, feature_converter_cls, seed,
File "/home/stephen/Desktop/t5x_final_test/t5x/t5x/utils.py", line 1387, in get_dataset_inner
ds = seqio.get_dataset(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/dataset_providers.py", line 1671, in get_dataset
ds = mixture_or_task.get_dataset(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/dataset_providers.py", line 1457, in get_dataset
datasets = [
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/dataset_providers.py", line 1458, in <listcomp>
task.get_dataset( # pylint:disable=g-complex-comprehension
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/dataset_providers.py", line 1209, in get_dataset
ds = self.preprocess_postcache(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/dataset_providers.py", line 1044, in preprocess_postcache
dataset = self._preprocess_dataset(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/dataset_providers.py", line 965, in _preprocess_dataset
dataset = prep_fn(dataset, **kwargs)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/preprocessors.py", line 83, in tokenize
return utils.map_over_dataset(fn=tokenize_fn)(dataset)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/utils.py", line 778, in wrapped_fn
return ds.map(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/data/ops/dataset_ops.py", line 2050, in map
return ParallelMapDataset(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/data/ops/dataset_ops.py", line 5284, in __init__
self._map_func = structured_function.StructuredFunctionWrapper(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/data/ops/structured_function.py", line 271, in __init__
self._function = fn_factory()
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/eager/function.py", line 2567, in get_concrete_function
graph_function = self._get_concrete_function_garbage_collected(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/eager/function.py", line 2533, in _get_concrete_function_garbage_collected
graph_function, _ = self._maybe_define_function(args, kwargs)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/eager/function.py", line 2711, in _maybe_define_function
graph_function = self._create_graph_function(args, kwargs)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/eager/function.py", line 2627, in _create_graph_function
func_graph_module.func_graph_from_py_func(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/framework/func_graph.py", line 1141, in func_graph_from_py_func
func_outputs = python_func(*func_args, **func_kwargs)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/data/ops/structured_function.py", line 248, in wrapped_fn
ret = wrapper_helper(*args)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/data/ops/structured_function.py", line 177, in wrapper_helper
ret = autograph.tf_convert(self._func, ag_ctx)(*nested_args)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 692, in wrapper
raise e.ag_error_metadata.to_exception(e)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 689, in wrapper
return converted_call(f, args, kwargs, options=options)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 439, in converted_call
result = converted_f(*effective_args, **kwargs)
File "/tmp/__autograph_generated_fileu9gu1w4n.py", line 8, in <lambda>
tf__lam = lambda arg: ag__.with_function_scope(lambda lscope: ag__.converted_call(fn, (arg,) + tuple(args), dict(**kargs), lscope), 'lscope', ag__.STD)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/core/function_wrappers.py", line 113, in with_function_scope
return thunk(scope)
File "/tmp/__autograph_generated_fileu9gu1w4n.py", line 8, in <lambda>
tf__lam = lambda arg: ag__.with_function_scope(lambda lscope: ag__.converted_call(fn, (arg,) + tuple(args), dict(**kargs), lscope), 'lscope', ag__.STD)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 352, in converted_call
return converted_call(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 439, in converted_call
result = converted_f(*effective_args, **kwargs)
File "/tmp/__autograph_generated_filezbhafqmt.py", line 113, in tf__tokenize_impl
ag__.for_stmt(ag__.converted_call(ag__.ld(features).items, (), None, fscope), None, loop_body, get_state_4, set_state_4, (), {'iterate_names': '(k, v)'})
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/operators/control_flow.py", line 449, in for_stmt
_py_for_stmt(iter_, extra_test, body, None, None)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/operators/control_flow.py", line 498, in _py_for_stmt
body(target)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/operators/control_flow.py", line 464, in protected_body
original_body(protected_iter)
File "/tmp/__autograph_generated_filezbhafqmt.py", line 105, in loop_body
ag__.if_stmt(ag__.ld(k) in ag__.ld(output_features), if_body_3, else_body_3, get_state_3, set_state_3, ('v',), 1)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/operators/control_flow.py", line 1341, in if_stmt
_py_if_stmt(cond, body, orelse)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/operators/control_flow.py", line 1394, in _py_if_stmt
return body() if cond else orelse()
File "/tmp/__autograph_generated_filezbhafqmt.py", line 63, in if_body_3
v = ag__.converted_call(ag__.ld(vocab).encode_tf, (ag__.ld(v),), None, fscope)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 441, in converted_call
result = converted_f(*effective_args)
File "/tmp/__autograph_generated_filef9jwq2ra.py", line 13, in tf__encode_tf
retval_ = ag__.converted_call(ag__.ld(self)._encode_tf, (ag__.ld(s),), None, fscope)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 441, in converted_call
result = converted_f(*effective_args)
File "/tmp/__autograph_generated_filezpl5g8b_.py", line 21, in tf___encode_tf
retval_ = ag__.converted_call(ag__.ld(self).tf_tokenizer.tokenize, (ag__.ld(s),), None, fscope)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 441, in converted_call
result = converted_f(*effective_args)
File "/tmp/__autograph_generated_filet9vre1mq.py", line 22, in tf__tokenize
input_tensor = ag__.converted_call(ag__.ld(ragged_tensor).convert_to_tensor_or_ragged_tensor, (ag__.ld(input),), None, fscope)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 377, in converted_call
return _call_unconverted(f, args, kwargs, options)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/autograph/impl/api.py", line 459, in _call_unconverted
return f(*args)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/ops/ragged/ragged_tensor.py", line 2683, in convert_to_tensor_or_ragged_tensor
return ops.convert_to_tensor_v2_with_dispatch(
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/util/traceback_utils.py", line 153, in error_handler
raise e.with_traceback(filtered_tb) from None
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow/python/framework/tensor_util.py", line 441, in make_tensor_proto
raise ValueError("None values not supported.")
ValueError: in user code:
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/utils.py", line 779, in None *
lambda arg: fn(arg, *args, **kargs)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/preprocessors.py", line 116, in tokenize_impl *
v = vocab.encode_tf(v)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/vocabularies.py", line 114, in encode_tf *
return self._encode_tf(s)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/vocabularies.py", line 413, in _encode_tf *
return self.tf_tokenizer.tokenize(s)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/tensorflow_text/python/ops/sentencepiece_tokenizer.py", line 133, in tokenize *
input_tensor = ragged_tensor.convert_to_tensor_or_ragged_tensor(input)
ValueError: None values not supported.
In call to configurable 'train' (<function train at 0x7f79d8db2280>)
This is how my task.py file looks
import functools
import seqio
import tensorflow as tf
import t5.data
from datasets import load_from_disk, load_dataset
from t5.data import postprocessors
from t5.data import preprocessors
from t5.evaluation import metrics
from seqio import FunctionDataSource, utils
TaskRegistry = seqio.TaskRegistry
vocabulary = seqio.SentencePieceVocabulary('gs://t5-data/vocabs/mc4.250000.100extra/sentencepiece.model', extra_ids=0)
DEFAULT_OUTPUT_FEATURES = {
"inputs": seqio.Feature(
vocabulary=vocabulary, add_eos=True,
required=False),
"targets": seqio.Feature(
vocabulary=vocabulary, add_eos=True)
}
def gen_dataset(split, shuffle=False, seed=None, column="text", dataset_path=None):
dataset = load_dataset(dataset_path, streaming=True, use_auth_token=True)
if shuffle:
if seed:
dataset = dataset.shuffle(seed=seed)
else:
dataset = dataset.shuffle()
while True:
for item in dataset[str(split)]:
yield item[column]
def dataset_fn(split, shuffle_files, seed=None, dataset_path=None):
return tf.data.Dataset.from_generator(
functools.partial(gen_dataset, split, shuffle_files, seed, dataset_path=dataset_path),
output_signature=tf.TensorSpec(shape=(), dtype=tf.string, name=dataset_path)
)
@utils.map_over_dataset
def target_to_key(x, key_map, target_key):
"""Assign the value from the dataset to target_key in key_map"""
return {**key_map, target_key: x}
# link to the mt5 sentencepiece tokenizer vocabulary
vocabulary = seqio.SentencePieceVocabulary('gs://t5-data/vocabs/mc4.250000.100extra/sentencepiece.model', extra_ids=0)
TaskRegistry.add(
"hindi_span_curruption",
source=seqio.FunctionDataSource(
dataset_fn=functools.partial(dataset_fn, dataset_path='StephennFernandes/ciil_mega_corpus_hindi'),
splits=("train", "validation"),
caching_permitted=False),
preprocessors=[
functools.partial(
target_to_key, key_map={
"inputs": None,
"targets": None,
}, target_key="targets"),
seqio.preprocessors.tokenize,
# seqio.CacheDatasetPlaceholder(),
preprocessors.span_corruption,
seqio.preprocessors.append_eos_after_trim,
],
output_features={"targets": DEFAULT_OUTPUT_FEATURES["targets"],"inputs": seqio.Feature(vocabulary=vocabulary,add_eos=True)},
metric_fns=[]
)
### similar multiple tasks exist for multiple languages. ###
seqio.MixtureRegistry.add(
"ciil_mix_3",
["assamese_span_curruption", "bengali_span_curruption",
"bhisnupuriya_span_curruption", "bodo_span_curruption",
"divehi_span_curruption", "dogri_span_curruption",
"english_span_curruption", "gujarati_span_curruption",
"hindi_span_curruption", "kannada_span_curruption",
"kashmiri_span_curruption", "konkani_span_curruption",
"maithili_span_curruption", "malayalam_span_curruption",
"manipuri_span_curruption", "marathi_span_curruption",
"nepali_span_curruption", "odia_span_curruption",
"panjabi_span_curruption", "sanskrit_span_curruption",
"tamil_span_curruption", "telugu_span_curruption",
"urdu_span_curruption" ],
default_rate=3
)
i further use the ciil_mix_3 mixture in my .gin file
this is how my .gin file looks
from __gin__ import dynamic_registration
import t5.data.mixtures
import __main__ as train_script
include 't5x/examples/t5/mt5/base.gin'
include 't5x/configs/runs/pretrain.gin'
import task
MIXTURE_OR_TASK_NAME = "ciil_mix_3"
TASK_FEATURE_LENGTHS = {"inputs": 512, "targets": 114}
TRAIN_STEPS = 100000
DROPOUT_RATE = 0.0
BATCH_SIZE = 32
train_script.train:
eval_period = 2000
I even further tried to work the same with byT5 and the same error occurs:
thr following is the error for byT5
ValueError: in user code:
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/utils.py", line 779, in None *
lambda arg: fn(arg, *args, **kargs)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/preprocessors.py", line 116, in tokenize_impl *
v = vocab.encode_tf(v)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/vocabularies.py", line 114, in encode_tf *
return self._encode_tf(s)
File "/home/stephen/anaconda3/lib/python3.9/site-packages/seqio/vocabularies.py", line 555, in _encode_tf *
tf_ids = tf.io.decode_raw(s, tf.uint8) + self._num_special_tokens
ValueError: Tried to convert 'bytes' to a tensor and failed. Error: None values not supported.
from t5x.
@hwchung27 @adarob as this issue seems a bit complex i have opened a new issue on this here #728
from t5x.
Related Issues (20)
- Missing import in t5x/checkpoints.py
- Regarding epochs and steps
- How to initialize a train state from a PyTree of parameters?
- Unable to install t5x on cloud TPU HOT 3
- download pile dataset is not working HOT 1
- Accessing data in requester pays bucket
- How can I save a model file?
- Checkpointing got stuck on Google Cloud TPU HOT 1
- can be t5x installed on windows ?
- [Question] Weight init for lm_head HOT 1
- Does fine tuning tasks re-initiate optimizer states?
- Installation Error
- `pw_checkpoints` undefined in `train.py`
- `load_t5x_checkpoint` Does not work with new checkpoints HOT 1
- module 'jax' has no attribute 'tree' HOT 2
- Flash attention
- E tensorflow/tsl/platform/cloud/curl_http_request.cc
- [train.py] UnboundLocalError: local variable 'checkpoint_manager' referenced before assignment HOT 2
- Error trying to import t5x HOT 5
- madlad often hallucinates while translating
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from t5x.