Alex Reichenbach

Python Bindings

2

I would like to use this application, but I've been struggling with writing python bindings for it. I have functioning code below, but it still uses files instead of memory...

Fixing bug as addressed in thread #79 . Run the tests before merging, as I haven't tested yet.

Convolution Layer error

9

On a deep cnn, I'm getting the error "Incompatible combination of dilation_rate with strides." The model works without problems within keras. Is padding not implemented? Is there a workaround? Architecture...

NaN Output

3

I have a CNN with architecture as shown below. ``` model = Sequential() model.add(Conv1D(32, 64, strides=2, padding='same', input_shape=(19200, 1)) model.add(BatchNormalization()) model.add(Activation('relu')) model.add(MaxPooling1D(pool_size=8)) model.add(Conv1D(64, 32, strides=2, padding='same')) model.add(BatchNormalization()) model.add(Activation('relu')) model.add(MaxPooling1D(pool_size=8)) model.add(Conv1D(128,...

Minor detail. But it'll make the README look prettier.

Build error

1

Upon building on an updated Mac with the given build string, I get the following traceback. ``` opencv_knn.cpp:32:15: error: expected ';' after expression cv::KNearest knn(trainingSet, labels); ^ ; opencv_knn.cpp:32:7: error:...

Behavior when missing quantization version

1

The problem happened below. Turns out it didn't include the "general.quantization_version" metadata. In the case that llama.cpp reads a file without a version, it assumes 2 (grep for the line...

Direct Policy Optimization

4

### 🚀 The feature, motivation, and pitch Hey all! Appreciate the work. Is there any word on whether DPO [(direct policy optimization)](https://arxiv.org/abs/2305.18290) will be integrated into the trlx library soon?...

Is it possible to set the data type from the cli interface? ``` python -m sglang.launch_server --model-path llava-v1.6-34b.Q8_0.gguf --tokenizer-path liuhaotian/llava-v1.6-34b-tokenizer --port 8888 --host 0.0.0.0 --enable-flashinfer --dtype bfloat16 ``` If not,...

mLlama load error with non-default vocabulary sizes

4

mllama has an unexplained offset in its tokenizer initialization. I assume it's to deal with an adding issue, but I'm unsure. If you add any special tokens to the model,...