issues
search
axolotl-ai-cloud
/
axolotl
Go ahead and axolotl questions
https://axolotl-ai-cloud.github.io/axolotl/
Apache License 2.0
7.48k
stars
808
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
numpy 2.1.0 was released, but incompatible with numba
#1849
winglian
closed
2 weeks ago
0
ensure that the bias is also in the correct dtype
#1848
winglian
closed
2 weeks ago
2
make the train_on_eos default to turn so all eos tokens are treated the same
#1847
winglian
closed
2 weeks ago
0
rename jamba example
#1846
xgal
closed
2 weeks ago
0
fix: prompt phi
#1845
JohanWork
closed
2 weeks ago
0
fix: phi system prompt
#1844
JohanWork
closed
2 weeks ago
2
feat: add jamba chat_template
#1843
xgal
closed
2 weeks ago
0
Model saving issue after training
#1842
gothaleshubham
opened
2 weeks ago
1
pretrain: fix with sample_packing=false
#1841
tmm1
closed
2 weeks ago
0
examples: fix tiny-llama pretrain yml syntax
#1840
tmm1
closed
2 weeks ago
0
docs: minor syntax highlight fix
#1839
tmm1
closed
2 weeks ago
0
ORPO results in `Cannot flatten integer dtype tensors`
#1838
maziyarpanahi
opened
2 weeks ago
3
fix: dont change quant storage dtype in case of fsdp
#1837
xgal
closed
2 weeks ago
0
New changes in PEFT breaks FSDP (RLHF)
#1836
maziyarpanahi
closed
2 weeks ago
12
fix so inference can be run against quantized models without adapters
#1834
winglian
opened
2 weeks ago
0
examples: Fix config llama3
#1833
JohanWork
opened
2 weeks ago
1
inst chat jinja template does not match prompt format used while training with `conversation: mistral`
#1832
nyxkrage
opened
2 weeks ago
0
Edge case for local dataset loading if its a folder
#1830
ccdv-ai
opened
2 weeks ago
0
Pass `trust_remote_code` to `load_dataset(...)` for `datasets>=2.20.0`
#1829
ccdv-ai
opened
2 weeks ago
0
optionally save the final FSDP model as a sharded state dict
#1828
winglian
closed
2 weeks ago
0
add validation to prevent 8bit lora finetuning on H100s
#1827
winglian
closed
2 weeks ago
2
bitsandbytes==0.43.3 can't be installed on mac
#1826
happylolonly
closed
3 weeks ago
1
fix: parse model_kwargs
#1825
NanoCode012
closed
3 weeks ago
0
fix: parse eager_attention from cfg
#1824
NanoCode012
closed
3 weeks ago
0
bump hf dependencies
#1823
winglian
closed
3 weeks ago
0
Warn if we override the chat template in the tokenizer config
#1822
chiwanpark
closed
1 week ago
4
update sklearn versrion, torch compile env vars, don't worry about failure on preprocess load model
#1821
winglian
closed
3 weeks ago
0
update tinyllama to use final instead of checkpoints [skip ci]
#1820
winglian
closed
4 weeks ago
0
autoawq 0.2.6 causing conflicting dependencies
#1819
Mr-Jeffery
opened
4 weeks ago
4
fix the incorrect `max_length` for chat template
#1818
chiwanpark
closed
4 weeks ago
0
fix z3 leaf configuration when not using lists
#1817
winglian
closed
4 weeks ago
0
Support LGAI-EXAONE model
#1816
shing100
closed
15 hours ago
0
Attempt to run multigpu in PR CI for now to ensure it works
#1815
winglian
closed
4 weeks ago
0
skip no commit to main on ci
#1814
winglian
closed
1 month ago
0
Mistral Nemo 12B training CUDA Out of memory only when enabling EVAL. On 2x3090Ti FSDP.
#1813
Nero10578
opened
1 month ago
3
update peft and transformers
#1811
winglian
closed
1 month ago
0
remove un-necessary zero-first guard as it's already called in a parent fn
#1810
winglian
closed
1 month ago
0
set z3 leaf for deepseek v2
#1809
winglian
closed
1 month ago
0
logging improvements
#1808
winglian
closed
1 month ago
0
A wrapper of RESTful API - Web server on the top of Axolotl
#1807
ulhaqi12
opened
1 month ago
2
Fix colab example notebook
#1805
srib
closed
1 month ago
0
There is no model saved in the checkpoint/final when lora_tuning llama3
#1804
leoozy
closed
1 month ago
0
One cycle lr
#1803
winglian
closed
1 month ago
0
`undefined symbol: cget_col_row_stats` when finetune in the axolotl-cloud docker container
#1802
chengchengpei
closed
1 month ago
2
fix roles to train defaults and make logging less verbose
#1801
winglian
closed
1 month ago
0
Can't preprocess dataset using meta-llama/Meta-Llama-3.1-8B model
#1800
ohmeow
closed
1 month ago
3
Zamba2AttentionDecoderLayer.forward() takes from 4 to 10 positional arguments but 11 were given
#1799
lucyknada
opened
1 month ago
2
publish axolotl images without extras in the tag name
#1798
winglian
closed
1 month ago
0
update test and main/nightly builds
#1797
winglian
closed
1 month ago
0
use 12.4.1 instead of 12.4 [skip-ci]
#1796
winglian
closed
1 month ago
0
Previous
Next