kohya-ss / kohya-ss/sd-scripts
WARNING Unsuppored keys for sd-scripts !
- Dominant language
- Python
- Stars
- 7.2k
- Forks
- 1.2k
- Avg merge
- 11m
- Merged PRs (30d)
- 2
Description
Thank you very much for your work!
I tried to integrate accelerated lora(Hyper-SD/Hyper-FLUX.1-dev-8steps-lora.safetensors) into flux_dev, but the following error occurred:
python networks/convert_flux_lora.py
WARNING Unsuppored keys for sd-scripts: dict_keys(['transformer.context_embedder.lora_A.weight', 'transformer.context_embedder.lora_B.weight', convert_flux_lora.py:238
'transformer.norm_out.linear.lora_A.weight', 'transformer.norm_out.linear.lora_B.weight', 'transformer.proj_out.lora_A.weight', 'transformer.proj_out.lora_B.weight',
'transformer.time_text_embed.guidance_embedder.linear_1.lora_A.weight', 'transformer.time_text_embed.guidance_embedder.linear_1.lora_B.weight',
'transformer.time_text_embed.guidance_embedder.linear_2.lora_A.weight', 'transformer.time_text_embed.guidance_embedder.linear_2.lora_B.weight',
'transformer.time_text_embed.text_embedder.linear_1.lora_A.weight', 'transformer.time_text_embed.text_embedder.linear_1.lora_B.weight',
'transformer.time_text_embed.text_embedder.linear_2.lora_A.weight', 'transformer.time_text_embed.text_embedder.linear_2.lora_B.weight',
'transformer.time_text_embed.timestep_embedder.linear_1.lora_A.weight', 'transformer.time_text_embed.timestep_embedder.linear_1.lora_B.weight',
'transformer.time_text_embed.timestep_embedder.linear_2.lora_A.weight', 'transformer.time_text_embed.timestep_embedder.linear_2.lora_B.weight',
'transformer.x_embedder.lora_A.weight', 'transformer.x_embedder.lora_B.weight'])
Contributor guide
No contributing guide indexed for this repository
Research direction
Start in networks/convert_flux_lora.py around line 238, where the warning lists unsupported sd-scripts keys. Trace how the listed Hyper-SD/Hyper-FLUX LoRA keys are handled, and consider the work complete when the relevant keys are supported without the warning during conversion.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 30/100