Skip to content

Commit 0f5b01b

Browse files
committed
dumb dockerfile test, add preliminary docs
1 parent 43fb8ef commit 0f5b01b

File tree

2 files changed

+18
-3
lines changed

2 files changed

+18
-3
lines changed

Dockerfile

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -177,9 +177,6 @@ RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
177177
RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
178178
PATH=$PATH:/opt/conda/bin make -C backend/python/transformers \
179179
; fi
180-
RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
181-
PATH=$PATH:/opt/conda/bin make -C backend/python/transformers-musicgen \
182-
; fi
183180
RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
184181
PATH=$PATH:/opt/conda/bin make -C backend/python/vall-e-x \
185182
; fi
@@ -192,6 +189,10 @@ RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
192189
RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
193190
PATH=$PATH:/opt/conda/bin make -C backend/python/petals \
194191
; fi
192+
RUN if [ "${IMAGE_TYPE}" = "extras" ]; then \
193+
pwd; \
194+
PATH=$PATH:/opt/conda/bin make -C backend/python/transformers-musicgen \
195+
; fi
195196

196197
# Define the health check command
197198
HEALTHCHECK --interval=1m --timeout=10m --retries=10 \

docs/content/features/text-to-audio.md

Lines changed: 14 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -52,6 +52,20 @@ Note:
5252
- The model name is case sensitive.
5353
- LocalAI must be compiled with the `GO_TAGS=tts` flag.
5454

55+
LocalAI also has experimental support for `transformers-musicgen` for the generation of short musical compositions. Currently, this is implemented via the same requests used for text to speech:
56+
57+
```
58+
curl --request POST \
59+
--url http://localhost:8080/tts \
60+
--header 'Content-Type: application/json' \
61+
--data '{
62+
"backend": "transformers-musicgen",
63+
"model": "facebook/musicgen-medium",
64+
"input": "Cello Rave"
65+
}' | aplay```
66+
67+
Future versions of LocalAI will expose additional control over audio generation beyond the text prompt.
68+
5569
#### Configuration
5670
5771
Audio models can be configured via `YAML` files. This allows to configure specific setting for each backend. For instance, backends might be specifying a voice or supports voice cloning which must be specified in the configuration file.

0 commit comments

Comments
 (0)