Fanyn

> To avoid InputStream from buffering anything, you'll need to set maximumSize to 0. do not work. I had done the test that changed code as to ` try (FFmpegFrameGrabber...

> It might be the format that doesn't support streams too well, and when it detects it cannot seek, it starts buffering everything...? Have you tried with another input format...

https://github.com/tensorflow/addons/blob/master/tensorflow_addons/layers/crf.py#L184

``` ~/miniconda3/envs/chenhao-env/lib/python3.7/site-packages/keras_contrib/layers/crf.py in recursion(self, input_energy, mask, go_backwards, return_sequences, return_logZ, input_length) 514 515 if mask is not None: --> 516 mask2 = K.cast(K.concatenate([mask, K.zeros_like(mask[:, :1])], axis=1), 517 K.floatx()) 518 constants.append(mask2) ```...

for Keras 2.2.5 `--> 516 mask2 = K.cast(K.concatenate([mask, K.zeros_like(mask[:, :1])], axis=1), 517 K.floatx())` try change to mask2 = K.concatenate([K.cast(mask, K.floatx()), K.zeros_like(mask[:, :1])], axis=1)

@kdja90 https://github.com/keras-team/keras-contrib/blob/master/keras_contrib/layers/crf.py in this file.

I have the same situation. how do you solve it

``` from plyfile import PlyData import numpy as np from io import BytesIO def process_ply_to_splat(ply_file_path, splat_file_path): plydata = PlyData.read(ply_file_path) vert = plydata["vertex"] buffer = BytesIO() SH_C0 = 0.28209479177387814 for v...

你把Qwen/QwQ-32B 配置文件中的tokenizer_config.json中的 chat_template字段中的`\\n`移除掉，即可。

> I'll go check it out. In the vllm version 0.7.3: ``` if n_local_reader > 0: # for local readers, we will: # 1. create a shared memory ring buffer...