fix(vall-e-x): correctly install reqs in environment (#1377)

Signed-off-by: GitHub <noreply@github.com>
Co-authored-by: mudler <mudler@users.noreply.github.com>

Dockerfile

@@ -185,12 +185,6 @@ RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
 	PATH=$PATH:/opt/conda/bin make -C backend/python/petals \
     ; fi
 
-# Copy VALLE-X as it's not a real "lib"
-# TODO: this is wrong - we should copy the lib into the conda env path
-RUN if [ -d /usr/lib/vall-e-x ]; then \
-    cp -rfv /usr/lib/vall-e-x/* ./ ; \ 
-    fi
-
 # we also copy exllama libs over to resolve exllama import error
 # TODO: check if this is still needed
 RUN if [ -d /usr/local/lib/python3.9/dist-packages/exllama ]; then \

Makefile

@@ -8,7 +8,7 @@ GOLLAMA_VERSION?=aeba71ee842819da681ea537e78846dc75949ac0
 
 GOLLAMA_STABLE_VERSION?=50cee7712066d9e38306eccadcfbb44ea87df4b7
 
-CPPLLAMA_VERSION?=1f5cd83275fabb43f2ae92c30033b384a3eb37b4
+CPPLLAMA_VERSION?=5a7d3125e7c24f223659b7f0b7aa7736986e92c0
 
 # gpt4all version
 GPT4ALL_REPO?=https://github.com/nomic-ai/gpt4all
@@ -22,7 +22,7 @@ RWKV_REPO?=https://github.com/donomii/go-rwkv.cpp
 RWKV_VERSION?=c898cd0f62df8f2a7830e53d1d513bef4f6f792b
 
 # whisper.cpp version
-WHISPER_CPP_VERSION?=85ed71aaec8e0612a84c0b67804bde75aa75a273
+WHISPER_CPP_VERSION?=e369243ebd24c8a14201f6b4280bccbb7b6a7df3
 
 # bert.cpp version
 BERT_VERSION?=6abe312cded14042f6b7c3cd8edf082713334a4d
@@ -85,11 +85,13 @@ endif
 
 ifeq ($(BUILD_TYPE),openblas)
 	CGO_LDFLAGS+=-lopenblas
+	export WHISPER_OPENBLAS=1
 endif
 
 ifeq ($(BUILD_TYPE),cublas)
 	CGO_LDFLAGS+=-lcublas -lcudart -L$(CUDA_LIBPATH)
 	export LLAMA_CUBLAS=1
+	export WHISPER_CUBLAS=1
 endif
 
 ifeq ($(BUILD_TYPE),hipblas)
@@ -98,6 +100,7 @@ ifeq ($(BUILD_TYPE),hipblas)
 	export CC=$(ROCM_HOME)/llvm/bin/clang
 	# llama-ggml has no hipblas support, so override it here.
 	export STABLE_BUILD_TYPE=
+	export WHISPER_HIPBLAS=1
 	GPU_TARGETS ?= gfx900,gfx90a,gfx1030,gfx1031,gfx1100
 	AMDGPU_TARGETS ?= "$(GPU_TARGETS)"
 	CMAKE_ARGS+=-DLLAMA_HIPBLAS=ON -DAMDGPU_TARGETS="$(AMDGPU_TARGETS)" -DGPU_TARGETS="$(GPU_TARGETS)"
@@ -107,10 +110,12 @@ endif
 ifeq ($(BUILD_TYPE),metal)
 	CGO_LDFLAGS+=-framework Foundation -framework Metal -framework MetalKit -framework MetalPerformanceShaders
 	export LLAMA_METAL=1
+	export WHISPER_METAL=1
 endif
 
 ifeq ($(BUILD_TYPE),clblas)
 	CGO_LDFLAGS+=-lOpenCL -lclblast
+	export WHISPER_CLBLAST=1
 endif
 
 # glibc-static or glibc-devel-static required
@@ -234,6 +239,7 @@ replace:
 	$(GOCMD) mod edit -replace github.com/go-skynet/go-ggml-transformers.cpp=$(shell pwd)/sources/go-ggml-transformers
 	$(GOCMD) mod edit -replace github.com/donomii/go-rwkv.cpp=$(shell pwd)/sources/go-rwkv
 	$(GOCMD) mod edit -replace github.com/ggerganov/whisper.cpp=$(shell pwd)/sources/whisper.cpp
+	$(GOCMD) mod edit -replace github.com/ggerganov/whisper.cpp/bindings/go=$(shell pwd)/sources/whisper.cpp/bindings/go
 	$(GOCMD) mod edit -replace github.com/go-skynet/go-bert.cpp=$(shell pwd)/sources/go-bert
 	$(GOCMD) mod edit -replace github.com/mudler/go-stable-diffusion=$(shell pwd)/sources/go-stable-diffusion
 	$(GOCMD) mod edit -replace github.com/mudler/go-piper=$(shell pwd)/sources/go-piper

backend/python/vall-e-x/Makefile

@@ -3,6 +3,7 @@ ttsvalle:
 	@echo "Creating virtual environment..."
 	@conda env create --name ttsvalle --file ttsvalle.yml
 	@echo "Virtual environment created."
+	bash install.sh
 
 .PHONY: run
 run:

backend/python/vall-e-x/install.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+
+##
+## A bash script installs the required dependencies of VALL-E-X and prepares the environment
+export PATH=$PATH:/opt/conda/bin
+
+# Activate conda environment
+source activate ttsvalle
+
+echo $CONDA_PREFIX
+
+git clone https://github.com/Plachtaa/VALL-E-X.git $CONDA_PREFIX/vall-e-x && pushd $CONDA_PREFIX/vall-e-x && pip install -r requirements.txt && popd
+
+cp -rfv $CONDA_PREFIX/vall-e-x/* ./

backend/python/vall-e-x/run.sh

@@ -10,4 +10,4 @@ source activate ttsvalle
 # get the directory where the bash script is located
 DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
 
-python $DIR/ttvalle.py $@
+python $DIR/ttsvalle.py $@

docs/content/advanced/_index.en.md

@@ -354,3 +354,13 @@ docker run --env REBUILD=true localai
 # Option 2: set within an env file
 docker run --env-file .env localai
 ```
+
+### Build only a single backend
+
+You can control the backends that are built by setting the `GRPC_BACKENDS` environment variable. For instance, to build only the `llama-cpp` backend only:
+
+```bash
+make GRPC_BACKENDS=backend-assets/grpc/llama-cpp build
+```
+
+By default, all the backends are built.

docs/content/advanced/fine-tuning.md

@@ -11,6 +11,8 @@ Section under construction
 
 This section covers how to fine-tune a language model for text generation and consume it in LocalAI.
 
+[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/mudler/LocalAI/blob/master/examples/e2e-fine-tuning/notebook.ipynb)
+
 ## Requirements
 
 For this example you will need at least a 12GB VRAM of GPU and a Linux box.

examples/e2e-fine-tuning/README.md

@@ -2,7 +2,7 @@ This is an example of fine-tuning a LLM model to use with [LocalAI](https://gith
 
 Specifically, this example shows how to use [axolotl](https://github.com/OpenAccess-AI-Collective/axolotl) to fine-tune a LLM model to consume with LocalAI as a `gguf` model.
 
-A notebook is provided that currently works on _very small_ datasets on Google colab on the free instance. It is far from producing good models, but it gives a sense of how to use the code to use with a better dataset and configurations, and how to use the model produced with LocalAI.
+A notebook is provided that currently works on _very small_ datasets on Google colab on the free instance. It is far from producing good models, but it gives a sense of how to use the code to use with a better dataset and configurations, and how to use the model produced with LocalAI. [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/mudler/LocalAI/blob/master/examples/e2e-fine-tuning/notebook.ipynb)
 
 ## Requirements