lllyasviel/stable-diffusion-webui-forge
1
0
Fork 0
mirror of https://github.com/lllyasviel/stable-diffusion-webui-forge.git synced 2026-02-25 09:13:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

move codes

Browse Source
This commit is contained in:
layerdiffusion
2024-08-31 11:31:02 -07:00
parent 33963f2d19
commit 79b25a8235
1 changed files with 7 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
backend/memory_management.py
View File

@@ -610,7 +610,8 @@ def load_models_gpu(models, memory_required=0, hard_memory_preservation=0):
global vram_state
execution_start_time = time.perf_counter()
extra_mem = max(minimum_inference_memory(), memory_required + hard_memory_preservation)
memory_to_free = max(minimum_inference_memory(), memory_required) + hard_memory_preservation
memory_for_inference = minimum_inference_memory() + hard_memory_preservation
models_to_load = []
models_already_loaded = []
@@ -628,7 +629,7 @@ def load_models_gpu(models, memory_required=0, hard_memory_preservation=0):
devs = set(map(lambda a: a.device, models_already_loaded))
for d in devs:
if d != torch.device("cpu"):
free_memory(extra_mem, d, models_already_loaded)
free_memory(memory_to_free, d, models_already_loaded)
moving_time = time.perf_counter() - execution_start_time
if moving_time > 0.1:
@@ -646,7 +647,7 @@ def load_models_gpu(models, memory_required=0, hard_memory_preservation=0):
for device in total_memory_required:
if device != torch.device("cpu"):
free_memory(total_memory_required[device] * 1.3 + extra_mem, device, models_already_loaded)
free_memory(total_memory_required[device] * 1.3 + memory_to_free, device, models_already_loaded)
for loaded_model in models_to_load:
model = loaded_model.model
@@ -661,16 +662,14 @@ def load_models_gpu(models, memory_required=0, hard_memory_preservation=0):
if lowvram_available and (vram_set_state == VRAMState.LOW_VRAM or vram_set_state == VRAMState.NORMAL_VRAM):
model_require = loaded_model.exclusive_memory
previously_loaded = loaded_model.inclusive_memory
current_free_mem = get_free_memory(torch_dev)
inference_memory = minimum_inference_memory() + hard_memory_preservation
estimated_remaining_memory = current_free_mem - model_require - inference_memory
estimated_remaining_memory = current_free_mem - model_require - memory_for_inference
print(f"[Memory Management] Target: {loaded_model.model.model.__class__.__name__}, Free GPU: {current_free_mem / (1024 * 1024):.2f} MB, Model Require: {model_require / (1024 * 1024):.2f} MB, Previously Loaded: {previously_loaded / (1024 * 1024):.2f} MB, Inference Require: {inference_memory / (1024 * 1024):.2f} MB, Remaining: {estimated_remaining_memory / (1024 * 1024):.2f} MB, ", end="")
print(f"[Memory Management] Target: {loaded_model.model.model.__class__.__name__}, Free GPU: {current_free_mem / (1024 * 1024):.2f} MB, Model Require: {model_require / (1024 * 1024):.2f} MB, Previously Loaded: {previously_loaded / (1024 * 1024):.2f} MB, Inference Require: {memory_for_inference / (1024 * 1024):.2f} MB, Remaining: {estimated_remaining_memory / (1024 * 1024):.2f} MB, ", end="")
if estimated_remaining_memory < 0:
vram_set_state = VRAMState.LOW_VRAM
model_gpu_memory_when_using_cpu_swap = compute_model_gpu_memory_when_using_cpu_swap(current_free_mem, inference_memory)
model_gpu_memory_when_using_cpu_swap = compute_model_gpu_memory_when_using_cpu_swap(current_free_mem, memory_for_inference)
if previously_loaded > 0:
model_gpu_memory_when_using_cpu_swap = previously_loaded