hi,
I'm having a few issues after loading the none.
if I load a non-Quantops FLUX1 or 2 using regular or KJ diffusion loader. I have the error message. Load the Quantops models works using QuantOps node.
No issues with Z_IMAGE non-Quantops diffusion using KJ. only flux. I haven't tested WAN
Removing the ComfyUI-QuantOps node everything works normal.
I have a 3090 Card
"triton.compiler.errors.CompilationError: at 1:0:
def dequantize_fp8_kernel_tl(
^
ValueError("type fp8e4nv not supported in this architecture. The supported fp8 dtypes are ('fp8e4b15', 'fp8e5')")
# ComfyUI Error Report
## Error Details
- **Node ID:** 61
- **Node Type:** KSampler
- **Exception Type:** triton.compiler.errors.CompilationError
- **Exception Message:** triton.compiler.errors.CompilationError: at 1:0:
def dequantize_fp8_kernel_tl(
^
ValueError("type fp8e4nv not supported in this architecture. The supported fp8 dtypes are ('fp8e4b15', 'fp8e5')")
## Stack Trace
File "/home/comfy/ComfyUI/execution.py", line 525, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
hi,
I'm having a few issues after loading the none.
if I load a non-Quantops FLUX1 or 2 using regular or KJ diffusion loader. I have the error message. Load the Quantops models works using QuantOps node.
No issues with Z_IMAGE non-Quantops diffusion using KJ. only flux. I haven't tested WAN
Removing the ComfyUI-QuantOps node everything works normal.
I have a 3090 Card
File "/home/comfy/ComfyUI/execution.py", line 525, in execute