Issues
- 14
thanks (it's 10x faster than JAX)!
#25 opened by Birch-san - 6
Execution stops with message "killed"
#16 opened by diapason-consulting - 3
Branches + Versions for maintainability
#35 opened by tataganesh - 0
Add Docker compatibility
#37 opened by lucaspar - 2
BusError only with mega model
#10 opened by amin-nejad - 1
cannot download datasets after login
#18 opened by pcunninghamCD - 9
ValueError: Unpack failed: incomplete input
#4 opened by jchia - 4
Incompatible with Python 3.10
#31 opened by matthewfranglen - 10
Recent commit results in higher VRAM usage
#39 opened by EeveeTea - 13
Text Tokenizer is fragmenting words
#32 opened by Kreevoz - 2
WARNING:absl:No GPU/TPU found, falling back to CPU
#22 opened by kenorb - 2
- 1
- 12
- 1
ExtraData: unpack(b) received extra data.
#34 opened by phatdoyle - 2
setup.sh fails due to possible CRLF issue
#33 opened by Filadeus - 3
High memory usage of torch compared to flax
#11 opened by phhusson - 5
--mega / is_mega raises TypeError: lax.dynamic_update_slice requires arguments to have the same dtypes, got float32, float16.
#15 opened by neilobremski - 3
- 4
Undeclared requirement - wandb api key
#7 opened by mikewarot - 1
OMP: Error #15: Initializing libiomp5.dylib, but found libomp.dylib already initialized.
#12 opened by letalvoj - 1
- 3
TypeError: lax.dynamic_update_slice requires arguments to have the same dtypes, got float16, float32
#13 opened by papul - 4
TypeError: lax.dynamic_update_slice requires arguments to have the same dtypes, got float32, float16.
#8 opened by andreafdaf - 6
Incorrect dtypes error with the Mega model
#2 opened by BenEaston - 5