Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
Spaces:
Dovakiins
/
qwerrwe
like
0
Build error
App
Files
Files
Community
Fetching metadata from the HF Docker repository...
d339beb
qwerrwe
/
src
/
axolotl
/
utils
100 contributors
History:
383 commits
Nanobit
fix(tokenizer): handle fast tokenizer properly for bos/eos (#914)
fde091c
unverified
about 1 year ago
samplers
don't train if eval split is too small (#873)
about 1 year ago
__init__.py
Safe
0 Bytes
WIP large refactor to make finetune script a little more manageable (#3)
almost 2 years ago
bench.py
Safe
2.06 kB
misc fixes to add gptq tests (#621)
over 1 year ago
callbacks.py
Safe
21.8 kB
feature: loss watchdog for terminating training runs that are failing (#899)
about 1 year ago
collators.py
6.34 kB
multipack w batch sampler (#795)
over 1 year ago
config.py
15.8 kB
Support device_map=sequential & max_memory config parameters (#903)
about 1 year ago
data.py
Safe
30.8 kB
fix: revert local dir dataset load (#878)
about 1 year ago
dict.py
Safe
326 Bytes
fix DefaultDict.__or__
over 1 year ago
distributed.py
Safe
7.07 kB
cleanup verbosity a bit
over 1 year ago
models.py
23.2 kB
fix(tokenizer): handle fast tokenizer properly for bos/eos (#914)
about 1 year ago
schedulers.py
Safe
3.58 kB
add axolotl trainer and quadratic warmup
over 1 year ago
tokenization.py
Safe
1.43 kB
add a latest tag for regular axolotl image, cleanup extraneous print statement (#746)
over 1 year ago
trainer.py
9.96 kB
Determine FSDP/deepspeed settings on device select. (#883)
about 1 year ago
wandb_.py
Safe
609 Bytes
Feat(wandb): Refactor to be more flexible (#767)
about 1 year ago