whisper.cpp/Makefile

#
# Audio samples
#

.PHONY: build
build:
	cmake -B build
	cmake --build build --config Release

# download a few audio samples into folder "./samples":
.PHONY: samples
samples:
	@echo "Downloading samples..."
	@mkdir -p samples
	@wget --quiet --show-progress -O samples/gb0.ogg https://upload.wikimedia.org/wikipedia/commons/2/22/George_W._Bush%27s_weekly_radio_address_%28November_1%2C_2008%29.oga
	@wget --quiet --show-progress -O samples/gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg
	@wget --quiet --show-progress -O samples/hp0.ogg https://upload.wikimedia.org/wikipedia/en/d/d4/En.henryfphillips.ogg
	@wget --quiet --show-progress -O samples/mm1.wav https://cdn.openai.com/whisper/draft-20220913a/micro-machines.wav
	@wget --quiet --show-progress -O samples/a13.mp3 https://upload.wikimedia.org/wikipedia/commons/transcoded/6/6f/Apollo13-wehaveaproblem.ogg/Apollo13-wehaveaproblem.ogg.mp3
	@wget --quiet --show-progress -O samples/diffusion2023-07-03.flac https://archive.org/download/diffusion2023-07-03/diffusion2023-07-03.flac
	@echo "Converting to 16-bit WAV ..."
	@ffmpeg -loglevel -0 -y -i samples/gb0.ogg -ar 16000 -ac 1 -c:a pcm_s16le samples/gb0.wav
	@ffmpeg -loglevel -0 -y -i samples/gb1.ogg -ar 16000 -ac 1 -c:a pcm_s16le samples/gb1.wav
	@ffmpeg -loglevel -0 -y -i samples/hp0.ogg -ar 16000 -ac 1 -c:a pcm_s16le samples/hp0.wav
	@rm samples/*.ogg
	@ffmpeg -loglevel -0 -y -i samples/mm1.wav -ar 16000 -ac 1 -c:a pcm_s16le samples/mm0.wav
	@rm samples/mm1.wav
	@ffmpeg -loglevel -0 -y -i samples/a13.mp3 -ar 16000 -ac 1 -c:a pcm_s16le -ss 00:00:00 -to 00:00:30 samples/a13.wav
	@rm samples/a13.mp3
	@ffmpeg -loglevel -0 -y -i samples/diffusion2023-07-03.flac -ar 16000 -ac 1 -c:a pcm_s16le samples/diffusion2023-07-03.wav
	@rm samples/diffusion2023-07-03.flac

#
# Models
#

# if not already downloaded, the following targets download the specified model and
# runs it on all samples in the folder "./samples":

.PHONY: tiny.en
.PHONY: tiny
.PHONY: base.en
.PHONY: base
.PHONY: small.en
.PHONY: small
.PHONY: medium.en
.PHONY: medium
.PHONY: large-v1
.PHONY: large-v2
.PHONY: large-v3
.PHONY: large-v3-turbo

tiny.en tiny base.en base small.en small medium.en medium large-v1 large-v2 large-v3 large-v3-turbo:
	bash ./models/download-ggml-model.sh $@
	cmake -B build
	cmake --build build --config Release
	@echo ""
	@echo "==============================================="
	@echo "Running $@ on all samples in ./samples ..."
	@echo "==============================================="
	@echo ""
	@for f in samples/*.wav; do \
		echo "----------------------------------------------" ; \
		echo "[+] Running $@ on $$f ... (run 'ffplay $$f' to listen)" ; \
	    echo "----------------------------------------------" ; \
		echo "" ; \
		./build/bin/main -m models/ggml-$@.bin -f $$f ; \
		echo "" ; \
	done
wip : improve makefile 2022-10-05 17:41:35 +00:00			`#`
			`# Audio samples`
			`#`

scripts : update to new build system 2024-12-09 09:30:16 +00:00			`.PHONY: build`
			`build:`
			`cmake -B build`
			`cmake --build build --config Release`

Update README.md and simplify usage 2022-09-26 06:36:51 +00:00			`# download a few audio samples into folder "./samples":`
Initial release 2022-09-25 18:23:15 +00:00			`.PHONY: samples`
			`samples:`
			`@echo "Downloading samples..."`
Create README.md 2022-09-25 19:35:26 +00:00			`@mkdir -p samples`
Initial release 2022-09-25 18:23:15 +00:00			`@wget --quiet --show-progress -O samples/gb0.ogg https://upload.wikimedia.org/wikipedia/commons/2/22/George_W._Bush%27s_weekly_radio_address_%28November_1%2C_2008%29.oga`
			`@wget --quiet --show-progress -O samples/gb1.ogg https://upload.wikimedia.org/wikipedia/commons/1/1f/George_W_Bush_Columbia_FINAL.ogg`
			`@wget --quiet --show-progress -O samples/hp0.ogg https://upload.wikimedia.org/wikipedia/en/d/d4/En.henryfphillips.ogg`
Reduce memory usage even more + better sampling - The encode/decode memory buffers are now reused - If the 30-sec segment goes for too long without a timestamp token, we force one. Improves transcription for large model - Stereo support - Add "micro-machines.wav" sample 2022-09-30 16:33:09 +00:00			`@wget --quiet --show-progress -O samples/mm1.wav https://cdn.openai.com/whisper/draft-20220913a/micro-machines.wav`
whisper : support speaker segmentation (local diarization) of mono audio via tinydiarize (#1058) * add HuggingFace mirror to download ggml model * support tdrz via simple hack overriding solm tokens * fix incorrect translate/transcribe token_ids that are not static const * add apollo 13 sample for tdrz demo * render [SPEAKER TURN] consistently in all terminal output using vocab.id_to_token * extend whisper_segment with speaker_turn_next field and save in json output * fix failing go build * slipped in some python syntax whoops * whisper : finalize tinydiarize support (add flag + fixes) * whisper : tdrz support for word-level timestamps (respect max_len) * java : try to fix tests after adding tdrz_enable flag * main : remove TODO leftover * java : fix params order list after adding "tdrz_enable" * whisper : fix solm and add nosp token * main : print tinydiarize help --------- Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> 2023-07-04 06:45:00 +00:00			`@wget --quiet --show-progress -O samples/a13.mp3 https://upload.wikimedia.org/wikipedia/commons/transcoded/6/6f/Apollo13-wehaveaproblem.ogg/Apollo13-wehaveaproblem.ogg.mp3`
samples : add a larger (30min) sample (#1092) Co-authored-by: Vadim Peretokin <vadim.peretokin@carasent.com> 2023-07-25 16:00:45 +00:00			`@wget --quiet --show-progress -O samples/diffusion2023-07-03.flac https://archive.org/download/diffusion2023-07-03/diffusion2023-07-03.flac`
Initial release 2022-09-25 18:23:15 +00:00			`@echo "Converting to 16-bit WAV ..."`
			`@ffmpeg -loglevel -0 -y -i samples/gb0.ogg -ar 16000 -ac 1 -c:a pcm_s16le samples/gb0.wav`
			`@ffmpeg -loglevel -0 -y -i samples/gb1.ogg -ar 16000 -ac 1 -c:a pcm_s16le samples/gb1.wav`
			`@ffmpeg -loglevel -0 -y -i samples/hp0.ogg -ar 16000 -ac 1 -c:a pcm_s16le samples/hp0.wav`
whisper : support speaker segmentation (local diarization) of mono audio via tinydiarize (#1058) * add HuggingFace mirror to download ggml model * support tdrz via simple hack overriding solm tokens * fix incorrect translate/transcribe token_ids that are not static const * add apollo 13 sample for tdrz demo * render [SPEAKER TURN] consistently in all terminal output using vocab.id_to_token * extend whisper_segment with speaker_turn_next field and save in json output * fix failing go build * slipped in some python syntax whoops * whisper : finalize tinydiarize support (add flag + fixes) * whisper : tdrz support for word-level timestamps (respect max_len) * java : try to fix tests after adding tdrz_enable flag * main : remove TODO leftover * java : fix params order list after adding "tdrz_enable" * whisper : fix solm and add nosp token * main : print tinydiarize help --------- Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> 2023-07-04 06:45:00 +00:00			`@rm samples/*.ogg`
Reduce memory usage even more + better sampling - The encode/decode memory buffers are now reused - If the 30-sec segment goes for too long without a timestamp token, we force one. Improves transcription for large model - Stereo support - Add "micro-machines.wav" sample 2022-09-30 16:33:09 +00:00			`@ffmpeg -loglevel -0 -y -i samples/mm1.wav -ar 16000 -ac 1 -c:a pcm_s16le samples/mm0.wav`
			`@rm samples/mm1.wav`
whisper : support speaker segmentation (local diarization) of mono audio via tinydiarize (#1058) * add HuggingFace mirror to download ggml model * support tdrz via simple hack overriding solm tokens * fix incorrect translate/transcribe token_ids that are not static const * add apollo 13 sample for tdrz demo * render [SPEAKER TURN] consistently in all terminal output using vocab.id_to_token * extend whisper_segment with speaker_turn_next field and save in json output * fix failing go build * slipped in some python syntax whoops * whisper : finalize tinydiarize support (add flag + fixes) * whisper : tdrz support for word-level timestamps (respect max_len) * java : try to fix tests after adding tdrz_enable flag * main : remove TODO leftover * java : fix params order list after adding "tdrz_enable" * whisper : fix solm and add nosp token * main : print tinydiarize help --------- Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> 2023-07-04 06:45:00 +00:00			`@ffmpeg -loglevel -0 -y -i samples/a13.mp3 -ar 16000 -ac 1 -c:a pcm_s16le -ss 00:00:00 -to 00:00:30 samples/a13.wav`
			`@rm samples/a13.mp3`
samples : add a larger (30min) sample (#1092) Co-authored-by: Vadim Peretokin <vadim.peretokin@carasent.com> 2023-07-25 16:00:45 +00:00			`@ffmpeg -loglevel -0 -y -i samples/diffusion2023-07-03.flac -ar 16000 -ac 1 -c:a pcm_s16le samples/diffusion2023-07-03.wav`
			`@rm samples/diffusion2023-07-03.flac`
Initial release 2022-09-25 18:23:15 +00:00
wip : improve makefile 2022-10-05 17:41:35 +00:00			`#`
			`# Models`
			`#`
Initial release 2022-09-25 18:23:15 +00:00
Update README.md and simplify usage 2022-09-26 06:36:51 +00:00			`# if not already downloaded, the following targets download the specified model and`
			`# runs it on all samples in the folder "./samples":`
Initial release 2022-09-25 18:23:15 +00:00
Update README.md and simplify usage 2022-09-26 06:36:51 +00:00			`.PHONY: tiny.en`
Flash + language support (ref #2) - Achieved big performance improvement + memory usage reduction - Can now translate / transcribe different languages 2022-09-28 17:46:05 +00:00			`.PHONY: tiny`
Update README.md and simplify usage 2022-09-26 06:36:51 +00:00			`.PHONY: base.en`
Flash + language support (ref #2) - Achieved big performance improvement + memory usage reduction - Can now translate / transcribe different languages 2022-09-28 17:46:05 +00:00			`.PHONY: base`
Initial release 2022-09-25 18:23:15 +00:00			`.PHONY: small.en`
Flash + language support (ref #2) - Achieved big performance improvement + memory usage reduction - Can now translate / transcribe different languages 2022-09-28 17:46:05 +00:00			`.PHONY: small`
			`.PHONY: medium.en`
			`.PHONY: medium`
models : add the new "large" model release by OpenAI The old "large" model is now renamed "large-v1". If you have been using it, make sure to rename it and download the new "large" model for best results. 2022-12-06 16:48:57 +00:00			`.PHONY: large-v1`
whisper : add support for large v3 (#1444) * whisper : add support for large v3 * bench : fix build + fix go bindings * bench : fix n_mels * models : update readme 2023-11-07 13:30:18 +00:00			`.PHONY: large-v2`
whisper : make large version explicit + fix data size units (#1493) 2023-11-15 17:42:25 +00:00			`.PHONY: large-v3`
whisper : add large-v3-turbo (#2440) 2024-10-01 12:57:06 +00:00			`.PHONY: large-v3-turbo`
Initial release 2022-09-25 18:23:15 +00:00
make : shim cmake 2024-12-06 13:34:53 +00:00			`tiny.en tiny base.en base small.en small medium.en medium large-v1 large-v2 large-v3 large-v3-turbo:`
refactoring : move main + stream in examples + other stuff 2022-10-25 16:13:08 +00:00			`bash ./models/download-ggml-model.sh $@`
make : shim cmake 2024-12-06 13:34:53 +00:00			`cmake -B build`
			`cmake --build build --config Release`
Create README.md 2022-09-25 19:35:26 +00:00			`@echo ""`
Initial release 2022-09-25 18:23:15 +00:00			`@echo "==============================================="`
Update README.md and simplify usage 2022-09-26 06:36:51 +00:00			`@echo "Running $@ on all samples in ./samples ..."`
Initial release 2022-09-25 18:23:15 +00:00			`@echo "==============================================="`
			`@echo ""`
			`@for f in samples/*.wav; do \`
			`echo "----------------------------------------------" ; \`
correct model name display on running samples 2022-11-25 03:24:08 +00:00			`echo "[+] Running $@ on $$f ... (run 'ffplay $$f' to listen)" ; \`
Initial release 2022-09-25 18:23:15 +00:00			`echo "----------------------------------------------" ; \`
			`echo "" ; \`
make : shim cmake 2024-12-06 13:34:53 +00:00			`./build/bin/main -m models/ggml-$@.bin -f $$f ; \`
Initial release 2022-09-25 18:23:15 +00:00			`echo "" ; \`
			`done`