Open Issues Need Help
View All on GitHub help wanted good first issue
mdnazib963/CoreFlow mdnazib963/CoreFlow 1.9 MB int8 SLM decoding 3,400+ tok/s on ONE CPU core — trained from scratch in pure PyTorch, served by a dependency-free C11 runtime. Cache-fit kernels, 99.4% parity with PyTorch.
Open project →
0
1.9 MB int8 SLM decoding 3,400+ tok/s on ONE CPU core — trained from scratch in pure PyTorch, served by a dependency-free C11 runtime. Cache-fit kernels, 99.4% parity with PyTorch.
Python
#avx2#c11#cache-optimization#cpu-inference#from-scratch#high-performance#inference#int8#language-model#llm#machine-learning#pytorch#quantization#slm#transformer
documentation good first issue
mdnazib963/CoreFlow mdnazib963/CoreFlow 1.9 MB int8 SLM decoding 3,400+ tok/s on ONE CPU core — trained from scratch in pure PyTorch, served by a dependency-free C11 runtime. Cache-fit kernels, 99.4% parity with PyTorch.
Open project →
0
1.9 MB int8 SLM decoding 3,400+ tok/s on ONE CPU core — trained from scratch in pure PyTorch, served by a dependency-free C11 runtime. Cache-fit kernels, 99.4% parity with PyTorch.
Python
#avx2#c11#cache-optimization#cpu-inference#from-scratch#high-performance#inference#int8#language-model#llm#machine-learning#pytorch#quantization#slm#transformer
good first issue
mdnazib963/CoreFlow mdnazib963/CoreFlow 1.9 MB int8 SLM decoding 3,400+ tok/s on ONE CPU core — trained from scratch in pure PyTorch, served by a dependency-free C11 runtime. Cache-fit kernels, 99.4% parity with PyTorch.
Open project →
0
1.9 MB int8 SLM decoding 3,400+ tok/s on ONE CPU core — trained from scratch in pure PyTorch, served by a dependency-free C11 runtime. Cache-fit kernels, 99.4% parity with PyTorch.
Python
#avx2#c11#cache-optimization#cpu-inference#from-scratch#high-performance#inference#int8#language-model#llm#machine-learning#pytorch#quantization#slm#transformer