issues
search
Ki6an
/
fastT5
⚡ boost inference speed of T5 models by 5x & reduce the model size by 3x.
Apache License 2.0
564
stars
72
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
transformers.generation_utils import GenerationMixin
#78
thakursc1
opened
3 months ago
2
Does UMT5 Supports?
#77
jhpassion0621
opened
7 months ago
2
TypeError: quantize_dynamic() got an unexpected keyword argument 'activation_type'
#76
ByUnal
opened
1 year ago
4
Dead kernel
#74
benzom
opened
1 year ago
0
TypeError: quantize_dynamic() got an unexpected keyword argument 'activation_type'
#73
32r81b
closed
1 year ago
1
Is fastT5 qunatization slower than pytorch dynamic quantization?
#72
parikshitsaikia1619
opened
1 year ago
0
compat transformers 4.28 & onnxruntime 1.14
#71
mzurek-teldoc
opened
1 year ago
0
Segmentation fault (core dumped)
#70
lisi-github
opened
1 year ago
5
GPU support
#69
p-christ
opened
1 year ago
0
flan-t5 support
#68
loretoparisi
opened
1 year ago
5
Issue with onnxruntime
#67
GenVr
opened
1 year ago
4
In the source code, the use of attention_mask is contradictory
#66
ghozn
opened
1 year ago
1
No such file or directory: '/content/encoder.embed_tokens.weight'
#65
alexfdo
opened
1 year ago
3
fastt5 not working with FastAPI gunicorn and docker
#64
kklivil
opened
1 year ago
1
Upgrade ONNX runtime
#63
dandiep
opened
1 year ago
1
M2M100 to ONNX
#61
sorgfresser
opened
2 years ago
0
Thank You and Demo Running in the Browser
#60
praeclarum
closed
2 years ago
2
Decoder's encoder_attention_mask input should be called decoder_attention_mask
#59
praeclarum
closed
2 years ago
1
Is it suit for other translation model like "Helsinki-NLP / opus-mt-en-de"?
#58
xyx361100238
closed
2 years ago
1
Mt5 model loading fails
#57
OriAlpha
opened
2 years ago
11
Fails to convert T0-3B
#56
redthing1
opened
2 years ago
2
Support for py3.10
#55
adityay121
closed
2 years ago
1
Not able to run onnx model of allen ai t5 small on GPU
#54
RKGhule
closed
2 years ago
1
Failed to create CUDAExecutionProvider
#53
ekaterinatretyak
closed
2 years ago
1
Unable to retrieve hidden_states
#52
vsoesanto
opened
2 years ago
2
Electra model loading into fastT5
#51
OriAlpha
closed
2 years ago
0
Conversion of decoder with past_key_values to float16.
#50
VikasOjha666
closed
2 years ago
5
Observing difference in outputs from decoder with IO bindings.
#49
VikasOjha666
closed
2 years ago
10
quantize_dynamic() got an unexpected keyword argument 'activation_type'
#48
katkardhanraj
closed
2 years ago
1
Support latest transformers
#46
Ki6an
closed
2 years ago
0
offline install error
#45
JingxinLee
opened
2 years ago
1
how to fix when convert model mT5 with max_length = 512
#44
batman-do
opened
2 years ago
3
How to fix when run convert model mT5
#43
batman-do
closed
2 years ago
1
GPU support for fastT5
#42
rupeshpoojary97
closed
2 years ago
3
Implemented the code for BART
#41
siddharth-sharma7
closed
2 years ago
0
Incompatible with transformers>=4.16
#40
nbroad1881
closed
2 years ago
1
Accuracy hit when using fastT5
#39
mbuotidem
opened
2 years ago
0
the method get_onnx_model() should not need the path to original model
#38
piegu
opened
2 years ago
2
Cuda support
#37
fmaoro
closed
2 years ago
3
Is fastT5 deepspeed compatible?
#36
jhillhouse92
closed
2 years ago
0
Explicit mt5 support
#35
sam-writer
closed
2 years ago
0
GPU Optimization
#34
ashissamal
opened
2 years ago
7
Errors when loading saved onnx files
#33
itaim
closed
2 years ago
10
small onnx optimizations
#32
sam-writer
closed
2 years ago
4
Where is model saved after it is converted to ONNX?
#31
Oxi84
closed
2 years ago
1
add `psutil` to deps in `setup.py`
#30
aseifert
closed
2 years ago
1
add custom models folder support in get_onnx_model
#29
WarrierRajeev
closed
2 years ago
3
get_onnx_model fails
#28
WarrierRajeev
closed
2 years ago
1
Updating fastT5?
#27
piegu
closed
2 years ago
6
default to 0 threads so ORT can choose
#26
sam-writer
closed
2 years ago
0
Next