Spaces:

DevSchretl
/

MicroGPT-gradio

Sleeping

App Files Files Community

MicroGPT-gradio / models

21.8 kB

Ctrl+K

Ctrl+K

2 contributors

History: 2 commits

Devdan Schretlen

Claude Opus 5 (1M context)

Add a KV cache so sampling costs one token's forward, not the whole context

d6d74b9 about 1 month ago

__init__.py

648 Bytes
Add a KV cache so sampling costs one token's forward, not the whole context about 1 month ago
gpt2.py

6.17 kB
Port the MicroGPT chat demo to a Gradio SDK Space on ZeroGPU 2 months ago
gpt3.py

14.3 kB
Add a KV cache so sampling costs one token's forward, not the whole context about 1 month ago
tokenizer.py

626 Bytes
Port the MicroGPT chat demo to a Gradio SDK Space on ZeroGPU 2 months ago