ef1cd3ac98
Replace Python-side GGUF download with llama-server's native -hf flag for HuggingFace repos. Add frontend variant picker so users can choose quantization (Q4_K_M, Q8_0, BF16, etc.) with file sizes. Fix vision detection via mmproj files instead of hardcoding is_vision=False.
30 lines
647 B
Python
30 lines
647 B
Python
"""
|
|
Model and LoRA configuration handling
|
|
"""
|
|
from .model_config import (
|
|
ModelConfig,
|
|
GgufVariantInfo,
|
|
is_vision_model,
|
|
scan_trained_loras,
|
|
load_model_defaults,
|
|
get_base_model_from_lora,
|
|
load_model_config,
|
|
list_gguf_variants,
|
|
MODEL_NAME_MAPPING,
|
|
UI_STATUS_INDICATORS,
|
|
)
|
|
from .checkpoints import scan_checkpoints
|
|
|
|
__all__ = [
|
|
'ModelConfig',
|
|
'GgufVariantInfo',
|
|
'is_vision_model',
|
|
'scan_trained_loras',
|
|
'load_model_defaults',
|
|
'get_base_model_from_lora',
|
|
'load_model_config',
|
|
'list_gguf_variants',
|
|
'MODEL_NAME_MAPPING',
|
|
'UI_STATUS_INDICATORS',
|
|
'scan_checkpoints',
|
|
]
|