Search results

Got curious about maximum memory usage so I used the "time -v" command. Not too bad ...peak utilization half a Gbyte running against 2 seconds speech sample. Without AVX emulation: Maximum resident set size (kbytes): 592528 With AVX emulation: Maximum resident set size (kbytes): 595620

Ouch! But yes, if I do that I'll go all the way and port the whole PBX over, recompile my open-source codecs, proper industrial SDCard with power-failure protection and wear-leveling, yada yada yada. But that's work and I really like running an Intel box so I'm going to wait & see if deepspeech...

...and so I tried the Intel AVX instruction emulator which I so gleefully touted above for my Z8350 ¢entOS 7 box. First I downloaded "sde-external-8.35.0-2019-03-11-lin.tar.bz2" from Intel and stuffed this into "/usr/local/sbin". Then I issued: $ cd /usr/local/sbin $ tar -xvjf sde*bz2 $ mv...

Most likely not tflite because since I can't find any traces of it on my box and I did not install with -tflite option, but I'll check this out in detail (both methods) as I have time in coming weeks. I'll also check out this official Intel AVX emulation which looks like it may be of help...

Okay grasshopper ;) that explains why I couldn't find any tflite instance on my system :crazy: If I'm understanding you correctly then I should set up a build environment on my little Z8350 sportster and then search the make file for a flag where I can disable use of AVX optimizations. I'll take...

Well, tflite requires a CPU which supports the AVX instructions. My Z8350 does not and so I get the same error. I suppose it may be possible to recompile tensorflow without reliance on AVX but will probably run slooooow.

Nice work ...GENTLEMEN! This needed to happen and it needed to happen oh so badly! Totally A_mazing, big hearty THANK YOU, and I hope it gets seriously noticed by the entire FreePBX community.

I found this (somewhat dated) tutorial about installation + test-run on Ubuntu and RasPi which might make a good initial read/overview. Unfortunately I'm totally overrun till mid February but will have a look at some point at bringing it up in CentOS. (I don't know much about this sort of stuff...

One thing about the deepspeech output is that it certainly is consistent (up above anyhow). If you can always count on it outputting the written forms of numerals then that's something people can work with programmatically (for IVR use and so on). Would be worse if sometimes it spit "seven" and...

I see it now, thanks. It's called "Messages Context" and for some hard to explain reason I had been looking for "Text Message Context" so I wasn't finding it! Yes, nice alternative to banging in the file :)

The trunk context is [sms-in] but the extension context is [sms-out] which are two very different things, the former called by the trunk and the latter called by the extensions. I have not tried without but in order to send SMS from an extension there needs to be somewhere to define which...

The new code does exactly the same thing in the same way, except for: (1) It's more configurable, allowing the same SMS message to go out to multiple extensions. (2) The old code might have failed to send out a message to some of the trunks IF you had more than one outbound trunk configured. A...

That's already covered by the VoIP.ms web page but you're right, I should have called it out specifically here too like this: Connectivity / Trunks / VoIPms202 ; Edit(icon) / PJSIP Settings / Advanced: Message Context: sms-in

Thanks for the comparison jerrm! There is s dirty little secret in speech recognition that I wish to illuminate here: Services like IBM and Google pass the raw output from the speech recognition engine through a contextual "semantics processing" engine which analyzes the grammar/context and then...

I've revised the original post to present a complex and a simple option. Both are tested working here. There are a LOT of unnecessary lines in it for debug purposes. This stuff is begging for an AGI script and message queuing when an extension is down and so on. I'm curious however why the...

This is a bit of a risky post and so I'm only going to do this once: https://pbxinaflash.com/community/threads/peepl-its-time-to-stop-whispering-to-the-clouds.24190/ It would require some development...

Have you guys tried OpenConnect? I switched from OpenVPN to OpenConnect 5 years ago and have never looked back. Never a problem and I'm using a split-tunnel on port 443 so it works EvErYwHeRe! :D https://en.wikipedia.org/wiki/OpenConnect http://ocserv.gitlab.io/www/index.html...

No it does not. The VoIP.ms example code is for a different scenario where your phone connects directly to their POP as an extension, in which case the extension number must be the same as your VoIP.ms DID number since you are using their code. When passing through your own PBX the scenario is...

Look: as of December, deepspeech (speech to text) software now has a mature API and it even runs on RasPI. https://github.com/mozilla/DeepSpeech/releases No more need for IncrediblePBX to punt our private words out to the gOOgle sponge and similar for recognition. This opens up opportunities to...

It's all good ...I just don't like bad information in too high a dosage because I feel compelled to correct it. It's like being barraged with many many questions but for no genuine benefit. Electricity is paid for in watt-hours because that's a measure of energy consumption, and this is why...