unslothai unsloth issues

unslothai / unsloth

Finetune Llama 3.2, Mistral, Phi, Qwen 2.5 & Gemma LLMs 2-5x faster with 80% less memory

https://unsloth.ai

Apache License 2.0

18.4k stars 1.29k forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

fix/load-checkpoint-add-new-tokens

#1225 Erland366 opened 3 weeks ago
3
OSError: could not get source code when loading a model using a for loop

#1224 daegonYu opened 3 weeks ago
4
Adding New Tokens

#1223 StrangePineAplle opened 3 weeks ago
6
FastLanguageModel.from_pretrained fails validate_repo_id in huggingface_hub

#1222 AndreBremer opened 3 weeks ago
4
Official Colab - unsloth/Llama-3.2-1B-Instruct-bnb-4bit randomly does not produce EOS tokens

#1221 jchook opened 3 weeks ago
6
Load And Unload Model Error: OSError: could not get source code

#1220 DaddyCodesAlot opened 3 weeks ago
4
Feat/all tmp

#1219 danielhanchen closed 3 weeks ago
1
Granite support

#1218 Datta0 opened 3 weeks ago
0
Potential bugfix in FlexAttention

#1217 AdityaKane2001 closed 3 weeks ago
2
Cross entropy for packing

#1216 fzyzcjy opened 3 weeks ago
2
Fail to load checkpoints trained with extended tokenizer

#1215 AbnetS opened 3 weeks ago
4
Error - 'OutOfMemoryError: CUDA out of memory.'

#1214 raghavendra-k-j opened 3 weeks ago
3
3B finetuned model - being Merged in to 7b Model, When saving to use in VLLM

#1213 pusapatiakhilraju closed 2 weeks ago
1
GGUF breaks

#1212 awesomecoolraj opened 3 weeks ago
2
Error saving PEFT adapter, re-loading model & adapter, and continuing to train

#1211 laura-burdick-sil closed 3 weeks ago
4
Continued Pre-Training Notebook not working with unsloth/Llama-3.2-1B-bnb-4bit

#1210 githomein opened 3 weeks ago
5
Please add the model: EleutherAI/polyglot-ko-5.8b

#1209 SabaPivot opened 3 weeks ago
1
Error `KeyError: 'layers.0.mlp.down_proj.weight'` when running Merged 4-bit Mistral Nemo in vLLM

#1208 josiah-redjade closed 3 weeks ago
3
Is there proper attention masking done when applying packing=true?

#1207 LostRuins opened 3 weeks ago
2
Installation for torch 2.5.0

#1206 Galaxy-Husky closed 3 weeks ago
1
Unable to use "unsloth/gemma-2b-bnb-4bit" model via vLLM

#1205 InderjeetVishnoi opened 3 weeks ago
6
merging w/ hacky gpu

#1204 Alex-Gurung closed 3 weeks ago
2
ORPO trainer not works after SFT

#1203 Romiroz opened 3 weeks ago
1
Question: How to fine tune an already finetuned model like NuExtract as a fine tune of Phi-3.5

#1202 KIC opened 3 weeks ago
2
Check if final_location is in /tmp in Kaggle environment

#1201 dendarrion closed 3 weeks ago
2
Fix/casting continue pretraining

#1200 Erland366 closed 3 weeks ago
3
Phi-3.5-mini generation becomes instable after 4096 tokens

#1199 NicolasSteen opened 3 weeks ago
1
Mistral Instruct v3 `sentencepiece_model.proto` error

#1198 CurtiusSimplus opened 3 weeks ago
25
Issue saving mistral-7b-instruct-v0.3-bnb-4bit to GGUF

#1197 Linguiniotta opened 3 weeks ago
6
Unable to execute FastLanguageModel.from_pretrained() with model unsloth/Llama-3.2-11B-Vision-Instruct-bnb-4bit on Google Colab

#1196 aeltorio closed 3 weeks ago
5
Bug fixes

#1195 danielhanchen closed 3 weeks ago
1
unsloth_train() does not work, shows more step than trainer.train()

#1194 Linguiniotta closed 3 weeks ago
2
Fix/phi-longrope

#1193 Erland366 closed 3 weeks ago
0
Train_on_completions cant handle eval_datasets as dictionary

#1192 R4ZZ3 opened 3 weeks ago
1
URGENT: unsloth saved lora adapter config not supported in VLLM

#1191 xinyudong93 closed 3 weeks ago
1
Errors with pip installation in Docker containers with torch 2.5

#1190 SyedA5688 closed 3 weeks ago
5
raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'") AttributeError: 'LongRopeRotaryEmbedding' object has no attribute 'long_cos_cached'. Did you mean: 'short_cos_cached'?

#1189 SnehaKumari14 opened 4 weeks ago
7
Cleanup upcast logs

#1188 Datta0 closed 3 weeks ago
0
pip install --upgrade --no-cache-dir unsloth BROKE CUDA packages. Inference slower.

#1187 pusapatiakhilraju opened 4 weeks ago
3
25% less mem and 10% faster training: Do not upcast lm_head and embedding to float32

#1186 Datta0 closed 4 weeks ago
0
ModuleNotFoundError : Failed to import transformers.models.falcon_mamba.configuration_falcon_mamba

#1185 CurtiusSimplus opened 4 weeks ago
2
RuntimeError: Expected out tensor to have dtype c10::BFloat16, but got float instead

#1184 Brightatkmitl opened 4 weeks ago
5
does unlsoth support freeze tunning

#1183 NathanaelTamirat opened 4 weeks ago
2
Fix 4.47 issue

#1182 danielhanchen closed 4 weeks ago
0
NameError: name 'Unpack' is not defined

#1181 CurtiusSimplus opened 4 weeks ago
12
fix/transformers-unpack

#1180 Erland366 closed 4 weeks ago
2
Can't import unsloth when both the latest version of unsloth and transformers are installed

#1179 lossflow opened 4 weeks ago
7
DPO, ORPO - grad accumulation fix

#1178 danielhanchen opened 4 weeks ago
0
Fix DPO, ORPO

#1177 danielhanchen closed 4 weeks ago
2
Unsloth full finetune: Does the fast speed and small memory come with a cost of performance degrading or not?

#1176 fzyzcjy opened 4 weeks ago
6

Previous Next