issues
search
AI-Hypercomputer
/
JetStream
JetStream is a throughput and memory optimized engine for LLM inference on XLA devices, starting with TPUs (and GPUs in future -- PRs welcome).
Apache License 2.0
202
stars
26
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Remove jax dependencies in JetStream
#88
FanhaiLu1
opened
4 months ago
0
Add ray disaggregated serving support
#87
FanhaiLu1
closed
4 months ago
2
Bump requests from 2.31.0 to 2.32.0 in the pip group across 1 directory
#86
dependabot[bot]
closed
4 months ago
0
Add profiling support and update docs
#85
JoeZijunZhou
closed
4 months ago
0
Allow tokenizer to customize stop_tokens
#84
qihqi
closed
4 months ago
0
Update benchmark script to easily test llama-3
#83
bhavya01
closed
4 months ago
0
Decode Batch Percentage Metrics/Improved Scraping
#82
Bslabe123
closed
4 months ago
0
Unit test coverage cleanup
#81
JoeZijunZhou
closed
4 months ago
0
Fix pad_tokens function description
#80
FanhaiLu1
closed
4 months ago
0
Support I/O with text and token ids
#79
JoeZijunZhou
closed
4 months ago
2
Update JetStream grpc proto to support I/O with text and token ids
#78
JoeZijunZhou
closed
4 months ago
8
Add metadata metrics
#77
yeandy
closed
5 months ago
0
Accuracy eval mlperf
#76
jwyang-google
closed
5 months ago
0
fix accuracy evaluation to align with MLPerf LLaMA2-chat models.
#75
jwyang-google
closed
5 months ago
0
Release v0.2.1 with docs update
#74
JoeZijunZhou
closed
5 months ago
1
Bump tqdm from 4.66.1 to 4.66.3 in the pip group across 1 directory
#73
dependabot[bot]
closed
5 months ago
0
Release v0.2.1
#72
JoeZijunZhou
closed
5 months ago
0
Prometheus Metrics
#71
Bslabe123
closed
4 months ago
6
Add tokenize_and_pad function to backward compatible
#70
FanhaiLu1
closed
5 months ago
0
Create __init__.py in Jetstream/third_party
#69
bhavya01
closed
5 months ago
0
Prerequisite work for supporting disaggregation:
#68
zhihaoshan-google
closed
5 months ago
0
Support llama3 tokenizer
#67
bhavya01
closed
5 months ago
0
Support llama3 tokenizer
#66
bhavya01
closed
5 months ago
0
refactor slice_to_num_chips to adapt to Cloud config
#65
zhihaoshan-google
closed
5 months ago
0
Detokenize error
#64
yeandy
closed
5 months ago
2
Register IFRT proxy backend when proxy is defined in the jax_platforms
#63
zhihaoshan-google
closed
5 months ago
0
Fix float division by zero in benchmark
#62
FanhaiLu1
closed
5 months ago
0
float division by zero in benchmark
#61
FanhaiLu1
opened
5 months ago
2
Fix requester flag default value
#60
JoeZijunZhou
closed
5 months ago
0
Fix benchmark script type issue
#59
JoeZijunZhou
closed
5 months ago
2
Benchmark serving: Failed to connect to remote host
#58
yeandy
closed
5 months ago
1
Error with mutable list value in dataclass
#57
yeandy
closed
5 months ago
1
Update maxtext user guide
#56
JoeZijunZhou
closed
5 months ago
0
Add np padding support
#55
FanhaiLu1
closed
4 months ago
1
Add jax_padding support driver and server lib
#54
FanhaiLu1
closed
5 months ago
0
Add an abstract class for Tokenizer
#53
bhavya01
closed
5 months ago
2
Do nothing for nd array in copy_to_host_async
#52
FanhaiLu1
closed
5 months ago
0
Format token utils and test
#51
FanhaiLu1
closed
5 months ago
0
Update README.md to fix broken link
#50
charbull
closed
5 months ago
0
Add np padded token support
#49
FanhaiLu1
closed
5 months ago
2
llama3 hack
#48
qihqi
closed
4 months ago
0
Add missing Documentation
#47
FanhaiLu1
closed
5 months ago
0
CogVLM support
#46
BitPhinix
closed
5 months ago
1
Refactor jestream to allow different tokenizers
#45
qihqi
opened
5 months ago
1
Support on Huggingface transformers
#44
ImKeTT
closed
5 months ago
2
Update server host default value
#43
JoeZijunZhou
closed
5 months ago
0
Reformat benchmark metrics
#42
yeandy
closed
5 months ago
3
Refactor readme
#41
FanhaiLu1
closed
5 months ago
6
Align Tokenizer in JetStream
#40
JoeZijunZhou
closed
5 months ago
0
Bump idna from 3.6 to 3.7
#39
dependabot[bot]
closed
5 months ago
0
Previous
Next