Spaces:
Sleeping
Sleeping
Commit History
-
test(tts): checking whether the tts model is working correctly
CrazyMonkey065952f6 -
fix: optimize FastAPI + Qwen2.5-1.5B for CPU, reduce max_tokens, increase timeout
CrazyMonkey088284a4
-
perf: implement lazy loading to fix startup timeouts
CrazyMonkey0- Load model on first request instead of startup - Increase token limits and Gunicorn timeout - Add stop token for cleaner responses8f110eb -
build: use custom Docker base image from Docker Hub
CrazyMonkey0- Replace python:3.12 with crazymonkey00/llama-base:latest - Remove llama-cpp-python installation (already in base image) - Reduce build time from 40+ minutes to ~30 secondsfc8b522 -
feat(docker): install llama-cpp-python directly with OpenBLAS
CrazyMonkey05adcb69 -
fix(docker): Add python alpine for llama-cpp-python
CrazyMonkey00442de3 -
fix(requirements): adapting packages for Python 3.13
CrazyMonkey0f42f941 -
fix(docker): switch version python to 3.12
CrazyMonkey0482015f -
fix(docker): switch to prebuilt llama-cpp-python image on Debian
CrazyMonkey033c584f -
fix(docker): repair version dependecies
CrazyMonkey05d2827d -
fix(docker): adding a permanent version of llama
CrazyMonkey006eacd3 -
fix(docker): replacing python:3.12-slim with python:3.12-alpine
CrazyMonkey0ad141ba -
fix(docker): add --prefer-binary
CrazyMonkey0347b2fa -
fix(docker): add --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu
CrazyMonkey0ef2c272 -
fix(docker): replacing python:3.12-alpine with python:3.12-slim
CrazyMonkey05c256f5 -
fix(docker): replacing apt-get with apk
CrazyMonkey038fe107 -
fix(docker): replacing python:3.12-slim with python:3.12-alpine
CrazyMonkey0bf92c5c -
fix(docker): replace llama-cpp-python extra index with pinned cp312 linux wheel
CrazyMonkey02a2937c -
fix(docker): changing the option --only-binary=llama-cpp-python to --prefer-binary
CrazyMonkey08757f21
-
fix(docker): add build dependencies for llama-cpp-python
CrazyMonkey0c0cf206 -
fix(docker): resolve llama-cpp-python module import error
CrazyMonkey0fd4e818 -
chore(docker): reintroduce llama-cpp-python pre-built wheel for faster build
CrazyMonkey023187e2
-
fix(docker): increase gunicorn worker timeout to 120s
CrazyMonkey0101988f -
feat(nlp): reintroduce Qwen2.5-1.5B-Instruct model and migrate back to Transformers
CrazyMonkey094cf754 -
fix(llama): fixed a bug with building llama
CrazyMonkey019edf1b
-
fix(llama): clear cache after llama
CrazyMonkey0ca4e6e3 -
test(models): downloading models from transformers
CrazyMonkey0fe8b413 -
fix(lama.cpp): Add prebuilt lamma.cpp to docker
CrazyMonkey0ad5570a -
fix(nlp): update Llama loading to use from_pretrained()
CrazyMonkey0f7ec4f4 -
feat(python): Change Python version to 3.11-bullseye for llama-cpp-python prebuilt wheel
CrazyMonkey0f45e402 -
Fix(nlp): NLP model download and build fix
CrazyMonkey089865a6 -
Fix(docker): repair building llama.cpp
CrazyMonkey074de35c -
Fix(docker): start of the model when building the Docker image
CrazyMonkey09c29a0e -
Fix commad to run in dockerfile
CrazyMonkey00b8f94b -
Fix python version
CrazyMonkey0420a9d9 -
Fix dockerfile
CrazyMonkey0385c547 -
Initial APP
CrazyMonkey07eb3110