myhloli
|
e7649e432c
refactor: modularize PDF processing and enhance async support for VLM backend
|
hace 4 meses |
myhloli
|
ec3adde809
fix: remove unused import of ModelPath in vlm_analyze.py
|
hace 5 meses |
myhloli
|
72c946a9ec
fix: include model_path in key for backend model retrieval in vlm_analyze.py
|
hace 5 meses |
myhloli
|
20dcbd2164
fix: include server_url in model key for backend model retrieval in vlm_analyze.py
|
hace 5 meses |
myhloli
|
013ddc02b2
fix: refine model path condition for transformers and sglang-engine backends in vlm_analyze.py
|
hace 5 meses |
myhloli
|
8737ebb2e2
feat: remove model path parameter from vlm_doc_analyze and streamline model loading
|
hace 5 meses |
myhloli
|
76259a80ff
fix: update backend references from huggingface to transformers in client and predictor modules
|
hace 5 meses |
myhloli
|
ab93f856f5
refactor: add 'local' option for model source and comment out timing logs
|
hace 5 meses |
myhloli
|
44a9cf220a
refactor: enhance model source handling and improve environment variable setup
|
hace 5 meses |
myhloli
|
cbba27b4f5
refactor: reorganize project structure and update import paths
|
hace 5 meses |
Jin Zhen Jiang
|
8e55a52693
feat: add mineru-vlm backend.
|
hace 5 meses |