mirror of
https://github.com/Monadical-SAS/reflector.git
synced 2025-12-20 12:19:06 +00:00
chore: remove old non-working code (#484)
This commit is contained in:
GitHub
@@ -1,204 +0,0 @@
|
||||
import re
|
||||
from pathlib import Path
|
||||
from typing import Any, List
|
||||
|
||||
from jiwer import wer
|
||||
from Levenshtein import distance
|
||||
from pydantic import BaseModel, Field, field_validator
|
||||
from tqdm.auto import tqdm
|
||||
from whisper.normalizers import EnglishTextNormalizer
|
||||
|
||||
|
||||
class EvaluationResult(BaseModel):
|
||||
"""
|
||||
Result object of the model evaluation
|
||||
"""
|
||||
accuracy: float = Field(default=0.0)
|
||||
total_test_samples: int = Field(default=0)
|
||||
|
||||
|
||||
class EvaluationTestSample(BaseModel):
|
||||
"""
|
||||
Represents one test sample
|
||||
"""
|
||||
|
||||
reference_text: str
|
||||
predicted_text: str
|
||||
|
||||
def update(self, reference_text:str, predicted_text:str) -> None:
|
||||
self.reference_text = reference_text
|
||||
self.predicted_text = predicted_text
|
||||
|
||||
|
||||
class TestDatasetLoader(BaseModel):
|
||||
"""
|
||||
Test samples loader
|
||||
"""
|
||||
|
||||
test_dir: Path = Field(default=Path(__file__).parent)
|
||||
total_samples: int = Field(default=0)
|
||||
|
||||
@field_validator("test_dir")
|
||||
def validate_file_path(cls, path):
|
||||
"""
|
||||
Check the file path
|
||||
"""
|
||||
if not path.exists():
|
||||
raise ValueError("Path does not exist")
|
||||
return path
|
||||
|
||||
def _load_test_data(self) -> tuple[Path, Path]:
|
||||
"""
|
||||
Loader function to validate input files and generate samples
|
||||
"""
|
||||
PREDICTED_TEST_SAMPLES_DIR = self.test_dir / "predicted_texts"
|
||||
REFERENCE_TEST_SAMPLES_DIR = self.test_dir / "reference_texts"
|
||||
|
||||
for filename in PREDICTED_TEST_SAMPLES_DIR.iterdir():
|
||||
match = re.search(r"(\d+)\.txt$", filename.as_posix())
|
||||
if match:
|
||||
sample_id = match.group(1)
|
||||
pred_file_path = PREDICTED_TEST_SAMPLES_DIR / filename
|
||||
ref_file_name = "ref_sample_" + str(sample_id) + ".txt"
|
||||
ref_file_path = REFERENCE_TEST_SAMPLES_DIR / ref_file_name
|
||||
if ref_file_path.exists():
|
||||
self.total_samples += 1
|
||||
yield ref_file_path, pred_file_path
|
||||
|
||||
def __iter__(self) -> EvaluationTestSample:
|
||||
"""
|
||||
Iter method for the test loader
|
||||
"""
|
||||
for pred_file_path, ref_file_path in self._load_test_data():
|
||||
with open(pred_file_path, "r", encoding="utf-8") as file:
|
||||
pred_text = file.read()
|
||||
with open(ref_file_path, "r", encoding="utf-8") as file:
|
||||
ref_text = file.read()
|
||||
yield EvaluationTestSample(reference_text=ref_text, predicted_text=pred_text)
|
||||
|
||||
|
||||
class EvaluationConfig(BaseModel):
|
||||
"""
|
||||
Model for evaluation parameters
|
||||
"""
|
||||
insertion_penalty: int = Field(default=1)
|
||||
substitution_penalty: int = Field(default=1)
|
||||
deletion_penalty: int = Field(default=1)
|
||||
normalizer: Any = Field(default=EnglishTextNormalizer())
|
||||
test_directory: str = Field(default=str(Path(__file__).parent))
|
||||
|
||||
|
||||
class ModelEvaluator:
|
||||
"""
|
||||
Class that comprises all model evaluation related processes and methods
|
||||
"""
|
||||
|
||||
# The 2 popular methods of WER differ slightly. More dimensions of accuracy
|
||||
# will be added. For now, the average of these 2 will serve as the metric.
|
||||
WEIGHTED_WER_LEVENSHTEIN = 0.0
|
||||
WER_LEVENSHTEIN = []
|
||||
WEIGHTED_WER_JIWER = 0.0
|
||||
WER_JIWER = []
|
||||
|
||||
evaluation_result = EvaluationResult()
|
||||
test_dataset_loader = None
|
||||
evaluation_config = None
|
||||
|
||||
def __init__(self, **kwargs):
|
||||
self.evaluation_config = EvaluationConfig(**kwargs)
|
||||
self.test_dataset_loader = TestDatasetLoader(test_dir=self.evaluation_config.test_directory)
|
||||
|
||||
def __repr__(self):
|
||||
return f"ModelEvaluator({self.evaluation_config})"
|
||||
|
||||
def describe(self) -> dict:
|
||||
"""
|
||||
Returns the parameters defining the evaluator
|
||||
"""
|
||||
return self.evaluation_config.model_dump()
|
||||
|
||||
def _normalize(self, sample: EvaluationTestSample) -> None:
|
||||
"""
|
||||
Normalize both reference and predicted text
|
||||
"""
|
||||
sample.update(
|
||||
self.evaluation_config.normalizer(sample.reference_text),
|
||||
self.evaluation_config.normalizer(sample.predicted_text),
|
||||
)
|
||||
|
||||
def _calculate_wer(self, sample: EvaluationTestSample) -> float:
|
||||
"""
|
||||
Based on weights for (insert, delete, substitute), calculate
|
||||
the Word Error Rate
|
||||
"""
|
||||
levenshtein_distance = distance(
|
||||
s1=sample.reference_text,
|
||||
s2=sample.predicted_text,
|
||||
weights=(
|
||||
self.evaluation_config.insertion_penalty,
|
||||
self.evaluation_config.deletion_penalty,
|
||||
self.evaluation_config.substitution_penalty,
|
||||
),
|
||||
)
|
||||
wer = levenshtein_distance / len(sample.reference_text)
|
||||
return wer
|
||||
|
||||
def _calculate_wers(self) -> None:
|
||||
"""
|
||||
Compute WER
|
||||
"""
|
||||
for sample in tqdm(self.test_dataset_loader, desc="Evaluating"):
|
||||
self._normalize(sample)
|
||||
wer_item_l = {
|
||||
"wer": self._calculate_wer(sample),
|
||||
"no_of_words": len(sample.reference_text),
|
||||
}
|
||||
wer_item_j = {
|
||||
"wer": wer(sample.reference_text, sample.predicted_text),
|
||||
"no_of_words": len(sample.reference_text),
|
||||
}
|
||||
self.WER_LEVENSHTEIN.append(wer_item_l)
|
||||
self.WER_JIWER.append(wer_item_j)
|
||||
|
||||
def _calculate_weighted_wer(self, wers: List[float]) -> float:
|
||||
"""
|
||||
Calculate the weighted WER from WER
|
||||
"""
|
||||
total_wer = 0.0
|
||||
total_words = 0.0
|
||||
for item in wers:
|
||||
total_wer += item["no_of_words"] * item["wer"]
|
||||
total_words += item["no_of_words"]
|
||||
return total_wer / total_words
|
||||
|
||||
def _calculate_model_accuracy(self) -> None:
|
||||
"""
|
||||
Compute model accuracy
|
||||
"""
|
||||
self._calculate_wers()
|
||||
weighted_wer_levenshtein = self._calculate_weighted_wer(self.WER_LEVENSHTEIN)
|
||||
weighted_wer_jiwer = self._calculate_weighted_wer(self.WER_JIWER)
|
||||
|
||||
final_weighted_wer = (weighted_wer_levenshtein + weighted_wer_jiwer) / 2
|
||||
self.evaluation_result.accuracy = (1 - final_weighted_wer) * 100
|
||||
|
||||
def evaluate(self, recalculate: bool = False) -> EvaluationResult:
|
||||
"""
|
||||
Triggers the model evaluation
|
||||
"""
|
||||
if not self.evaluation_result.accuracy or recalculate:
|
||||
self._calculate_model_accuracy()
|
||||
return EvaluationResult(
|
||||
accuracy=self.evaluation_result.accuracy,
|
||||
total_test_samples=self.test_dataset_loader.total_samples
|
||||
)
|
||||
|
||||
|
||||
eval_config = {"insertion_penalty": 1, "deletion_penalty": 2, "substitution_penalty": 1}
|
||||
|
||||
evaluator = ModelEvaluator(**eval_config)
|
||||
evaluation = evaluator.evaluate()
|
||||
|
||||
print(evaluator)
|
||||
print(evaluation)
|
||||
print("Model accuracy : {:.2f} %".format(evaluation.accuracy))
|
||||
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
File diff suppressed because one or more lines are too long
File diff suppressed because it is too large
Load Diff
@@ -1,620 +0,0 @@
|
||||
Technologies ticker symbol w-e-l-l on
|
||||
|
||||
the TSX recently reported its 2023 q1
|
||||
|
||||
results beating the streets consensus
|
||||
|
||||
estimate for revenue and adjusted ebitda
|
||||
|
||||
and in a report issued this week Raymond
|
||||
|
||||
James analyst said quote we're impressed
|
||||
|
||||
by Wells capacity to drive powerful
|
||||
|
||||
growth across its diverse business units
|
||||
|
||||
in the absence of M A joining me today
|
||||
|
||||
is CEO Hamed chabazi to look at what's
|
||||
|
||||
next for well health good to see you sir
|
||||
|
||||
how are you great to see you Richard
|
||||
|
||||
thanks very much for having me great to
|
||||
|
||||
have you uh congratulations on your 17th
|
||||
|
||||
consecutive quarter of record Revenue
|
||||
|
||||
can you share some insights into what's
|
||||
|
||||
Driven these results historically and in
|
||||
|
||||
the past quarter as well
|
||||
|
||||
yeah thank you we we're very excited
|
||||
|
||||
about our uh q1 2023 results and as you
|
||||
|
||||
mentioned uh we've had a long you know
|
||||
|
||||
successful uh string of of uh you know
|
||||
|
||||
continued growth and record growth
|
||||
|
||||
um we also had accelerating organic
|
||||
|
||||
growth and I think um a big part of the
|
||||
|
||||
success of our franchise here is the
|
||||
|
||||
incredibly sticky and predictable
|
||||
|
||||
Revenue that we have you know well over
|
||||
|
||||
90 of our business is either highly
|
||||
|
||||
reoccurring as in uh the you know highly
|
||||
|
||||
predictable uh results of our two-sided
|
||||
|
||||
network of patients and providers or
|
||||
|
||||
truly recurring as in scheduled or
|
||||
|
||||
subscribed revenues and this allows us
|
||||
|
||||
to essentially make sure that that uh
|
||||
|
||||
you know we're on track it obviously you
|
||||
|
||||
know like any other business things
|
||||
|
||||
happen uh and sometimes it's hard to
|
||||
|
||||
meet those results but what's really
|
||||
|
||||
being unique about our platform is we do
|
||||
|
||||
have exposure to all kinds of different
|
||||
|
||||
aspects of healthcare you know we have
|
||||
|
||||
Prime primary care and Specialized Care
|
||||
|
||||
on both sides of the Border in the US
|
||||
|
||||
and Canada so we have exposure to
|
||||
|
||||
different types of business models we
|
||||
|
||||
have exposure to the U.S payer Network
|
||||
|
||||
which has higher per unit economics than
|
||||
|
||||
Canada and of course the stability and
|
||||
|
||||
uh and and sort of higher Fidelity uh
|
||||
|
||||
kind of Collections and revenue cycle
|
||||
|
||||
process that Canada has over the United
|
||||
|
||||
States where you don't have to kind of
|
||||
|
||||
deal with all of that uh at that payment
|
||||
|
||||
noise so just a lot of I think strength
|
||||
|
||||
built into the platform because of the
|
||||
|
||||
diversity of different Healthcare
|
||||
|
||||
businesses that we support
|
||||
|
||||
and uh where do you see Well's future
|
||||
|
||||
growth coming from which part of the
|
||||
|
||||
business uh excites you the most right
|
||||
|
||||
now yeah well look the centrifugal force
|
||||
|
||||
of well is the healthcare provider and
|
||||
|
||||
we exist to uh Tech enable and
|
||||
|
||||
ameliorate the business of that of that
|
||||
|
||||
Tech of that healthcare provider uh and
|
||||
|
||||
and and that's what we're laser focused
|
||||
|
||||
on and and what we're seeing is
|
||||
|
||||
providers not wanting to run businesses
|
||||
|
||||
anymore it's very simple and so we have
|
||||
|
||||
a digital platform and providers can
|
||||
|
||||
either acquire what they want and need
|
||||
|
||||
from our digital platform and implement
|
||||
|
||||
it themselves
|
||||
|
||||
or they can decide that they don't want
|
||||
|
||||
to run a business anymore they don't
|
||||
|
||||
want to configure and manage technology
|
||||
|
||||
which is becoming a bigger and bigger
|
||||
|
||||
part of their world every single day and
|
||||
|
||||
when we see what we've seen with that
|
||||
|
||||
Dynamic is that uh is that a lot of them
|
||||
|
||||
are now just wanting to work in a place
|
||||
|
||||
where where all the technology is
|
||||
|
||||
configured for them it's wrapped around
|
||||
|
||||
them and they have a competent operating
|
||||
|
||||
partner that is supporting the organ the
|
||||
|
||||
the practice uh and and taking care of
|
||||
|
||||
the front office in the back office so
|
||||
|
||||
that they can focus on providing care
|
||||
|
||||
this results in them seeing more
|
||||
|
||||
patients uh and and being happier
|
||||
|
||||
because you know they became doctors to
|
||||
|
||||
see patients not so they can manage uh
|
||||
|
||||
workers and and deal with HR issues and
|
||||
|
||||
deal with labs and all that kind of
|
||||
|
||||
stuff excellent and I know too that
|
||||
|
||||
Acquisitions have played a key role in
|
||||
|
||||
well can you share any insights into how
|
||||
|
||||
the Acquisitions fit into Wells growth
|
||||
|
||||
strategy
|
||||
|
||||
sure in in look in 2020 and 2021 we did
|
||||
|
||||
a lot of Acquisitions in 2022 we took a
|
||||
|
||||
bit of a breather and we've really
|
||||
|
||||
focused on integration and I think
|
||||
|
||||
that's one of the reasons why you saw
|
||||
|
||||
this accelerating organic growth we
|
||||
|
||||
really were able to demonstrate that we
|
||||
|
||||
could bring together the different
|
||||
|
||||
elements of our technology platform we
|
||||
|
||||
started to sell bundles we started to
|
||||
|
||||
really derive Synergy uh and activate uh
|
||||
|
||||
you know more sales as a result of
|
||||
|
||||
selling uh all the different products
|
||||
|
||||
and services with one voice with One
|
||||
|
||||
Vision uh so we made it easier for
|
||||
|
||||
providers to use their technology and I
|
||||
|
||||
think that was a big reason uh for our
|
||||
|
||||
growth now M A as you know where Capital
|
||||
|
||||
allocation company we're never far from
|
||||
|
||||
it and so we did continue to have you
|
||||
|
||||
know tuck-ins here and there and in fact
|
||||
|
||||
today uh we announced that we've
|
||||
|
||||
acquired uh the Alberta operations of uh
|
||||
|
||||
MCI one Health and other publicly traded
|
||||
|
||||
company uh who was looking to raise
|
||||
|
||||
funds to support their business we're
|
||||
|
||||
very pleased with with this acquisition
|
||||
|
||||
it just demonstrates our continued
|
||||
|
||||
discipline these are you know great
|
||||
|
||||
primary care clinics in in Canada right
|
||||
|
||||
in the greater Calgary area and uh uh
|
||||
|
||||
you know just allows us to grow our
|
||||
|
||||
footprint in Alberta which is an
|
||||
|
||||
important Province for us and it it's
|
||||
|
||||
it's if you look at the price if you
|
||||
|
||||
look at what we're getting uh you know
|
||||
|
||||
it's just demonstrative of our continued
|
||||
|
||||
uh discipline and just you know a few
|
||||
|
||||
days ago at our conference call I
|
||||
|
||||
mentioned uh that we had you know a
|
||||
|
||||
really strong lineup of Acquisitions uh
|
||||
|
||||
and you know they're starting to uh uh I
|
||||
|
||||
think uh come to fruition for us
|
||||
|
||||
a company on the grown-up question I you
|
||||
|
||||
recently announced a new AI investment
|
||||
|
||||
program last month what specific areas
|
||||
|
||||
of healthcare technology or AI are you
|
||||
|
||||
focusing on and what's the strategy when
|
||||
|
||||
it comes to AI
|
||||
|
||||
yes uh look AI as as I'm sure you're
|
||||
|
||||
aware is it's become you know really uh
|
||||
|
||||
an incredibly important topic in in all
|
||||
|
||||
aspects of of business and and you know
|
||||
|
||||
not just business socially as well
|
||||
|
||||
everyone's talking about uh this this
|
||||
|
||||
new breakthrough disruptive technology
|
||||
|
||||
the large language models and generative
|
||||
|
||||
AI
|
||||
|
||||
um I mean look AI uh has been about a 80
|
||||
|
||||
year old overnight success a lot of
|
||||
|
||||
people have been working on this for a
|
||||
|
||||
long time generative AI is just sort of
|
||||
|
||||
you know the culmination of a lot of
|
||||
|
||||
things coming together and working uh
|
||||
|
||||
but it is uncorked enormous uh
|
||||
|
||||
Innovation and and we think that um this
|
||||
|
||||
there's a very good news story about
|
||||
|
||||
this in healthcare particularly where we
|
||||
|
||||
were looking to look we were looking to
|
||||
|
||||
unlock uh the value of of the data that
|
||||
|
||||
that we all produce every single day
|
||||
|
||||
um as as humans and and so we've
|
||||
|
||||
established an AI investment program
|
||||
|
||||
because no one company can can tackle
|
||||
|
||||
all of these Innovations themselves and
|
||||
|
||||
what well has done too is it's taken a
|
||||
|
||||
very much an ecosystem approach by
|
||||
|
||||
establishing its apps.health Marketplace
|
||||
|
||||
and so we're very excited about not only
|
||||
|
||||
uh allocating Capital into promising
|
||||
|
||||
young AI companies that are focused on
|
||||
|
||||
digital health and solving Healthcare
|
||||
|
||||
problems but also giving them access to
|
||||
|
||||
um you know safely and securely to our
|
||||
|
||||
provider Network to our uh you know to
|
||||
|
||||
to our Outpatient Clinic Network which
|
||||
|
||||
is the largest owned and operated
|
||||
|
||||
Network in Canada by far uh so
|
||||
|
||||
um and and when these and it's it was
|
||||
|
||||
remarkable when we announced this
|
||||
|
||||
program we've had just in the in the
|
||||
|
||||
first uh week to 10 days we've had over
|
||||
|
||||
a hundred uh inbound prospects come in
|
||||
|
||||
uh that that wanted to you know
|
||||
|
||||
collaborate with us and again I don't
|
||||
|
||||
think that's necessarily for the money
|
||||
|
||||
you know we're saying we would invest a
|
||||
|
||||
minimum of a quarter of a million
|
||||
|
||||
dollars you know a lot of them will
|
||||
|
||||
likely be higher than a quarter of a
|
||||
|
||||
million dollars
|
||||
|
||||
so it's not life-changing money but but
|
||||
|
||||
our structural advantages and and and
|
||||
|
||||
the benefits that we have in the Well
|
||||
|
||||
Network those are extremely hard to come
|
||||
|
||||
by uh and I think and I think uh uh
|
||||
|
||||
you'll see us uh you know help some of
|
||||
|
||||
these companies uh succeed and they will
|
||||
|
||||
help us drive uh you know more
|
||||
|
||||
Innovation to that helps the provider
|
||||
|
||||
but speaking of this very interesting AI
|
||||
|
||||
I know your company just launched well
|
||||
|
||||
AI voice this is super interesting tell
|
||||
|
||||
me what it is and the impact it could
|
||||
|
||||
have on health care providers
|
||||
|
||||
yeah thanks for uh asking Richard our
|
||||
|
||||
providers uh are thrilled with this you
|
||||
|
||||
know we've we've had a number of of of
|
||||
|
||||
our own well providers testing this
|
||||
|
||||
technology and it it it really feels
|
||||
|
||||
like magic to them it's essentially an
|
||||
|
||||
ambient AI powered scribe so it's a it's
|
||||
|
||||
a service that with the consent of the
|
||||
|
||||
parties involved listens to the
|
||||
|
||||
conversation between a patient and
|
||||
|
||||
provider and then uh essentially
|
||||
|
||||
condenses that into a medically relevant
|
||||
|
||||
note for the chart files uh typically
|
||||
|
||||
that is a lengthy process a doctor has
|
||||
|
||||
to transcribe notes then review those
|
||||
|
||||
notes and make sure that uh a a a a
|
||||
|
||||
appropriate medically oriented and
|
||||
|
||||
structured node is is is uh prepared and
|
||||
|
||||
put into the chart and that could take
|
||||
|
||||
you know sometimes more than more time
|
||||
|
||||
than the actual consultation uh time and
|
||||
|
||||
so we believe that on average if it's
|
||||
|
||||
used regularly and consistently this can
|
||||
|
||||
give providers back at least a third of
|
||||
|
||||
their day
|
||||
|
||||
um and and it's it's just a game changer
|
||||
|
||||
uh and and uh we have now gone into
|
||||
|
||||
General release with this product it's
|
||||
|
||||
widely available in Canada uh it has
|
||||
|
||||
been integrated into our EMR which makes
|
||||
|
||||
it even more valuable tools like this
|
||||
|
||||
are going to start popping up but if
|
||||
|
||||
they're not integrated into your
|
||||
|
||||
practice management system then you have
|
||||
|
||||
to kind of have data in in more than one
|
||||
|
||||
place and and move that around a little
|
||||
|
||||
bit which which makes it a little bit
|
||||
|
||||
more difficult especially with HIPAA
|
||||
|
||||
requirements and and regulations so
|
||||
|
||||
again I think this is the first of many
|
||||
|
||||
types of different products and services
|
||||
|
||||
that allow doctors to place more
|
||||
|
||||
emphasis and focus on the patient
|
||||
|
||||
experience instead of having their head
|
||||
|
||||
in a laptop and looking at you once in a
|
||||
|
||||
while they'll be looking at you and
|
||||
|
||||
speaking to their practice management
|
||||
|
||||
system and I think this you know think
|
||||
|
||||
about it as Alexa for for our doctors uh
|
||||
|
||||
you know this this ability to speak uh
|
||||
|
||||
and and have you know uh you know Voice
|
||||
|
||||
driven AI assistant that does things
|
||||
|
||||
like this I think are going to be you
|
||||
|
||||
know incredibly helpful and valuable uh
|
||||
|
||||
for for healthcare providers
|
||||
|
||||
super fascinating I mean we're just
|
||||
|
||||
hearing you know more about AI maybe AI
|
||||
|
||||
for the first time but here you are with
|
||||
|
||||
a product already on the market in the
|
||||
|
||||
in the healthcare field that's going to
|
||||
|
||||
be pretty attractive to be out there uh
|
||||
|
||||
right ahead of many other people right
|
||||
|
||||
thank you Richard thanks for that
|
||||
|
||||
recognition that's been Our intention we
|
||||
|
||||
we want to demonstrate that we uh you
|
||||
|
||||
know that we're all in on ensuring that
|
||||
|
||||
technology that benefits providers uh is
|
||||
|
||||
is is accelerated and uh de-risked and
|
||||
|
||||
provided uh you know um in in a timely
|
||||
|
||||
way you know providers need this help we
|
||||
|
||||
we have a healthcare crisis in the
|
||||
|
||||
country that is generally characterized
|
||||
|
||||
as a as a lack of doctors and so imagine
|
||||
|
||||
if we can get our doctors to be 20 or 30
|
||||
|
||||
percent more productive through the use
|
||||
|
||||
of these types of tools well they're
|
||||
|
||||
going to just see more patience and and
|
||||
|
||||
that's going to help all of us and uh
|
||||
|
||||
and look if you step back Wells business
|
||||
|
||||
model is all about having exposure to
|
||||
|
||||
the success of doctors and doing our
|
||||
|
||||
best to help them be more successful
|
||||
|
||||
because we're in a revenue share
|
||||
|
||||
relationship with most of the doctors
|
||||
|
||||
that we work with and so this uh this is
|
||||
|
||||
good for the ecosystem it's great for
|
||||
|
||||
the provider and it's great for well as
|
||||
|
||||
well super fascinating I'm Ed shabazzi
|
||||
|
||||
CEO well Health Technologies ticker
|
||||
|
||||
w-e-l-l great to catch up again thank
|
||||
|
||||
you sir
|
||||
|
||||
thank you Richard appreciate you having
|
||||
|
||||
me
|
||||
|
||||
[Music]
|
||||
|
||||
thank you
|
||||
|
||||
@@ -1,970 +0,0 @@
|
||||
learning medicine is hard work osmosis
|
||||
|
||||
makes it easy it takes our lectures and
|
||||
|
||||
notes to create a personalized study
|
||||
|
||||
plan with exclusive videos practice
|
||||
|
||||
questions and flashcards and so much
|
||||
|
||||
more try it free today
|
||||
|
||||
in diabetes mellitus your body has
|
||||
|
||||
trouble moving glucose which is the type
|
||||
|
||||
of sugar from your blood into your cells
|
||||
|
||||
this leads to high levels of glucose in
|
||||
|
||||
your blood and not enough of it in your
|
||||
|
||||
cells and remember that your cells need
|
||||
|
||||
glucose as a source of energy so not
|
||||
|
||||
letting the glucose enter means that the
|
||||
|
||||
cells star for energy despite having
|
||||
|
||||
glucose right on their doorstep in
|
||||
|
||||
general the body controls how much
|
||||
|
||||
glucose is in the blood relative to how
|
||||
|
||||
much gets into the cells with two
|
||||
|
||||
hormones insulin and glucagon insulin is
|
||||
|
||||
used to reduce blood glucose levels and
|
||||
|
||||
glucagon is used to increase blood
|
||||
|
||||
glucose levels both of these hormones
|
||||
|
||||
are produced by clusters of cells in the
|
||||
|
||||
pancreas called islets of langerhans
|
||||
|
||||
insulin is secreted by beta cells in the
|
||||
|
||||
center of these islets and glucagon is
|
||||
|
||||
secreted by alpha cells in the periphery
|
||||
|
||||
of the islets insulin reduces the amount
|
||||
|
||||
of glucose in the blood by binding to
|
||||
|
||||
insulin receptors embedded in the cell
|
||||
|
||||
membrane of various insulin responsive
|
||||
|
||||
tissues like muscle cells in adipose
|
||||
|
||||
tissue when activated the insulin
|
||||
|
||||
receptors cause vesicles containing
|
||||
|
||||
glucose transporter that are inside the
|
||||
|
||||
cell to fuse with the cell membrane
|
||||
|
||||
allowing glucose to be transported into
|
||||
|
||||
the cell glucagon does exactly the
|
||||
|
||||
opposite it raises the blood glucose
|
||||
|
||||
levels by getting the liver to generate
|
||||
|
||||
new molecules of glucose from other
|
||||
|
||||
molecules and also break down glycogen
|
||||
|
||||
into glucose so that I can all get
|
||||
|
||||
dumped into the blood diabetes mellitus
|
||||
|
||||
is diagnosed when blood glucose levels
|
||||
|
||||
get too high and this is seen among 10
|
||||
|
||||
percent of the world population there
|
||||
|
||||
are two types of diabetes type 1 and
|
||||
|
||||
type 2 and the main difference between
|
||||
|
||||
them is the underlying mechanism that
|
||||
|
||||
causes the blood glucose levels to rise
|
||||
|
||||
about 10% of people with diabetes have
|
||||
|
||||
type 1 and the remaining 90% of people
|
||||
|
||||
with diabetes have type 2 let's start
|
||||
|
||||
with type 1 diabetes mellitus sometimes
|
||||
|
||||
just called type 1 diabetes in this
|
||||
|
||||
situation the body doesn't make enough
|
||||
|
||||
insulin the reason this happens is that
|
||||
|
||||
in type 1 diabetes there's a type 4
|
||||
|
||||
hypersensitivity response or a cell
|
||||
|
||||
mediated immune response where a
|
||||
|
||||
person's own T cells at
|
||||
|
||||
the pancreas as a quick review remember
|
||||
|
||||
that the immune system has T cells that
|
||||
|
||||
react to all sorts of antigens which are
|
||||
|
||||
usually small peptides polysaccharides
|
||||
|
||||
or lipids and that some of these
|
||||
|
||||
antigens are part of our own body cells
|
||||
|
||||
it doesn't make sense to allow T cells
|
||||
|
||||
that will attack our own cells to hang
|
||||
|
||||
around until there's this process to
|
||||
|
||||
eliminate them called self tolerance in
|
||||
|
||||
type 1 diabetes there's a genetic
|
||||
|
||||
abnormality that causes a loss of self
|
||||
|
||||
tolerance among T cells that
|
||||
|
||||
specifically target the beta cell
|
||||
|
||||
antigens losing self tolerance means
|
||||
|
||||
that these T cells are allowed to
|
||||
|
||||
recruit other immune cells and
|
||||
|
||||
coordinate an attack on these beta cells
|
||||
|
||||
losing beta cells means less insulin and
|
||||
|
||||
less insulin means that glucose piles up
|
||||
|
||||
in the blood because it can't enter the
|
||||
|
||||
body's cells one really important group
|
||||
|
||||
of genes involved in regulation of the
|
||||
|
||||
immune response is the human leukocyte
|
||||
|
||||
antigen system or HLA system even though
|
||||
|
||||
it's called a system it's basically this
|
||||
|
||||
group of genes on chromosome 6 that
|
||||
|
||||
encode the major histocompatibility
|
||||
|
||||
complex or MHC which is a protein that's
|
||||
|
||||
extremely important in helping the
|
||||
|
||||
immune system recognize foreign
|
||||
|
||||
molecules as well as maintaining self
|
||||
|
||||
tolerance MHC is like the serving
|
||||
|
||||
platter that antigens are presented to
|
||||
|
||||
the immune cells on interestingly people
|
||||
|
||||
with type 1 diabetes often have specific
|
||||
|
||||
HLA genes in common with each other one
|
||||
|
||||
called
|
||||
|
||||
HLA dr3 and another called HLA dr4 but
|
||||
|
||||
this is just a genetic clue right
|
||||
|
||||
because not everyone with HLA dr3 and
|
||||
|
||||
HLA dr4 develops diabetes in diabetes
|
||||
|
||||
mellitus type 1 destruction of beta
|
||||
|
||||
cells usually starts early in life but
|
||||
|
||||
sometimes up to 90% of the beta cells
|
||||
|
||||
are destroyed before symptoms crop up
|
||||
|
||||
for clinical symptoms of uncontrolled
|
||||
|
||||
diabetes that all sound similar our
|
||||
|
||||
polyphagia glycosuria polyuria and
|
||||
|
||||
polydipsia let's go through them one by
|
||||
|
||||
one even though there's a lot of glucose
|
||||
|
||||
in the blood it cannot get into the
|
||||
|
||||
cells which leaves cells starved for
|
||||
|
||||
energy so in response adipose tissue
|
||||
|
||||
starts breaking down fat called
|
||||
|
||||
lipolysis
|
||||
|
||||
and muscle tissue starts breaking down
|
||||
|
||||
proteins both of which results in weight
|
||||
|
||||
loss for someone with uncontrolled
|
||||
|
||||
diabetes this catabolic state leaves
|
||||
|
||||
people feeling hungry
|
||||
|
||||
also known as poly fascia Faiza means
|
||||
|
||||
eating and poly means a lot now with
|
||||
|
||||
high glucose levels that means that when
|
||||
|
||||
blood gets filtered through the kidneys
|
||||
|
||||
some of it starts to spill into the
|
||||
|
||||
urine called glycosuria glyco surfers to
|
||||
|
||||
glucose and urea the urine since glucose
|
||||
|
||||
is osmotically active water tends to
|
||||
|
||||
follow it resulting in an increase in
|
||||
|
||||
urination or polyuria poly again refers
|
||||
|
||||
to a lot and urea again refers to urine
|
||||
|
||||
finally because there's so much
|
||||
|
||||
urination people with uncontrolled
|
||||
|
||||
diabetes become dehydrated and thirsty
|
||||
|
||||
or polydipsia poly means a lot and dip
|
||||
|
||||
SIA means thirst even though people with
|
||||
|
||||
diabetes are not able to produce their
|
||||
|
||||
own insulin they can still respond to
|
||||
|
||||
insulin so treatment involves lifelong
|
||||
|
||||
insulin therapy to regulate their blood
|
||||
|
||||
glucose levels and basically enable
|
||||
|
||||
their cells to use glucose
|
||||
|
||||
one really serious complication with
|
||||
|
||||
type 1 diabetes is called diabetic
|
||||
|
||||
ketoacidosis or DKA to understand it
|
||||
|
||||
let's go back to the process of
|
||||
|
||||
lipolysis where fat is broken down into
|
||||
|
||||
free fatty acids after that happens the
|
||||
|
||||
liver turns the fatty acids into ketone
|
||||
|
||||
bodies like Osito acetic acid in beta
|
||||
|
||||
hydroxy butyrate acid a seed of acetic
|
||||
|
||||
acid is a keto acid because it has a
|
||||
|
||||
ketone group in a carboxylic acid group
|
||||
|
||||
beta hydroxy rhetoric acid on the other
|
||||
|
||||
hand even though it's still one of the
|
||||
|
||||
ketone bodies isn't technically a keto
|
||||
|
||||
acid since its ketone group has been
|
||||
|
||||
reduced to a hydroxyl group these ketone
|
||||
|
||||
bodies are important because they can be
|
||||
|
||||
used by cells for energy but they also
|
||||
|
||||
increase the acidity of the blood which
|
||||
|
||||
is why it's called ketoacidosis and the
|
||||
|
||||
blood becoming really acidic can have
|
||||
|
||||
major effects throughout the body
|
||||
|
||||
individuals can develop custom all
|
||||
|
||||
respiration which is a deep and labored
|
||||
|
||||
breathing as the body tries to move
|
||||
|
||||
carbon dioxide out of the blood in an
|
||||
|
||||
effort to reduce its acidity cells also
|
||||
|
||||
have a transporter that exchanges
|
||||
|
||||
hydrogen ions or protons for potassium
|
||||
|
||||
when the blood gets acidic it's by
|
||||
|
||||
definition loaded with protons that get
|
||||
|
||||
sent into cells while potassium gets
|
||||
|
||||
sent into the fluid outside cells
|
||||
|
||||
another thing to keep in mind is that in
|
||||
|
||||
addition to helping glucose enter cells
|
||||
|
||||
insulin stimulates the sodium potassium
|
||||
|
||||
ATPase --is which help potassium get
|
||||
|
||||
into the cells and so without insulin
|
||||
|
||||
more potassium stays in the fluid
|
||||
|
||||
outside cells both of these mechanisms
|
||||
|
||||
lead to increased potassium in the fluid
|
||||
|
||||
outside cells which quickly makes it
|
||||
|
||||
into the blood and causes hyperkalemia
|
||||
|
||||
the potassium is then excreted so over
|
||||
|
||||
time even though the blood potassium
|
||||
|
||||
levels remain high over all stores of
|
||||
|
||||
potassium in the body which include
|
||||
|
||||
potassium inside cells starts to run low
|
||||
|
||||
individuals will also have a high anion
|
||||
|
||||
gap which reflects a large difference in
|
||||
|
||||
the unmeasured negative and positive
|
||||
|
||||
ions in the serum largely due to the
|
||||
|
||||
build-up of ketoacids
|
||||
|
||||
diabetic ketoacidosis can happen even in
|
||||
|
||||
people who have already been diagnosed
|
||||
|
||||
with diabetes and currently have some
|
||||
|
||||
sort of insulin therapy
|
||||
|
||||
in states of stress like an infection
|
||||
|
||||
the body releases epinephrine which in
|
||||
|
||||
turn stimulates the release of glucagon
|
||||
|
||||
too much glucagon can tip the delicate
|
||||
|
||||
hormonal balance of glucagon and insulin
|
||||
|
||||
in favor of elevating blood sugars and
|
||||
|
||||
can lead to a cascade of events we just
|
||||
|
||||
described increased glucose in the blood
|
||||
|
||||
loss of glucose in the urine loss of
|
||||
|
||||
water dehydration and in parallel and
|
||||
|
||||
need for alternative energy generation
|
||||
|
||||
of ketone bodies and ketoacidosis
|
||||
|
||||
interestingly both ketone bodies break
|
||||
|
||||
down into acetone and escape as a gas by
|
||||
|
||||
getting breathed out the lungs which
|
||||
|
||||
gives us sweet fruity smell to a
|
||||
|
||||
person's breath in general though that's
|
||||
|
||||
the only sweet thing about this illness
|
||||
|
||||
which also causes nausea vomiting and if
|
||||
|
||||
severe mental status changes and acute
|
||||
|
||||
cerebral edema
|
||||
|
||||
treatment of a DKA episode involves
|
||||
|
||||
giving plenty of fluids which helps with
|
||||
|
||||
dehydration insulin which helps lower
|
||||
|
||||
blood glucose levels and replacement of
|
||||
|
||||
electrolytes like potassium all of which
|
||||
|
||||
help to reverse the acidosis now let's
|
||||
|
||||
switch gears and talk about type 2
|
||||
|
||||
diabetes which is where the body makes
|
||||
|
||||
insulin but the tissues don't respond as
|
||||
|
||||
well to it the exact reason why cells
|
||||
|
||||
don't respond isn't fully understood
|
||||
|
||||
essentially the body's providing the
|
||||
|
||||
normal amount of insulin but the cells
|
||||
|
||||
don't move their glucose transporters to
|
||||
|
||||
their membrane in response which
|
||||
|
||||
remember is needed for the glucose to
|
||||
|
||||
get into the cells these cells therefore
|
||||
|
||||
have insulin resistance some risk
|
||||
|
||||
factors for insulin resistance are
|
||||
|
||||
obesity lack of exercise and
|
||||
|
||||
hypertension the exact mechanisms are
|
||||
|
||||
still being explored for example in
|
||||
|
||||
excess of adipose tissue or fat is
|
||||
|
||||
thought to cause the release of free
|
||||
|
||||
fatty acids in so-called edible kinds
|
||||
|
||||
which are signaling molecules that can
|
||||
|
||||
cause inflammation which seems related
|
||||
|
||||
to insulin resistance
|
||||
|
||||
however many people that are obese are
|
||||
|
||||
not diabetic so genetic factors probably
|
||||
|
||||
play a major role as well we see this
|
||||
|
||||
when we look at twin studies as well
|
||||
|
||||
we're having a twin with type-2 diabetes
|
||||
|
||||
increases the risk of developing type 2
|
||||
|
||||
diabetes completely independently of
|
||||
|
||||
other environmental risk factors in type
|
||||
|
||||
2 diabetes since tissues don't respond
|
||||
|
||||
as well to normal levels of insulin the
|
||||
|
||||
body ends up producing more insulin in
|
||||
|
||||
order to get the same effect and move
|
||||
|
||||
glucose out of the blood
|
||||
|
||||
they do this through beta cell
|
||||
|
||||
hyperplasia an increased number of beta
|
||||
|
||||
cells and beta cell hypertrophy where
|
||||
|
||||
they actually grow in size all in this
|
||||
|
||||
attempt to pump out more insulin this
|
||||
|
||||
works for a while and by keeping insulin
|
||||
|
||||
levels higher than normal blood glucose
|
||||
|
||||
levels can be kept normal called normal
|
||||
|
||||
glycemia now along with insulin beta
|
||||
|
||||
cells also secrete islet amyloid
|
||||
|
||||
polypeptide or amylin so while beta
|
||||
|
||||
cells are cranking out insulin they also
|
||||
|
||||
secrete an increased amount of amylin
|
||||
|
||||
over time Emlyn builds up and aggregates
|
||||
|
||||
in the islets this beta cell
|
||||
|
||||
compensation though is not sustainable
|
||||
|
||||
and over time those maxed out beta cells
|
||||
|
||||
get exhausted and they become
|
||||
|
||||
dysfunctional and undergo hypo trophy
|
||||
|
||||
and get smaller as well as hypoplasia
|
||||
|
||||
and die off as beta cells are lost in
|
||||
|
||||
insulin levels decrease glucose levels
|
||||
|
||||
in the blood start to increase in
|
||||
|
||||
patients develop hyperglycemia which
|
||||
|
||||
leads to similar clinical signs that we
|
||||
|
||||
mentioned before like Paul aphasia
|
||||
|
||||
glycosuria polyuria polydipsia but
|
||||
|
||||
unlike type 1 diabetes there's generally
|
||||
|
||||
some circulating insulin in type 2
|
||||
|
||||
diabetes from the beta cells that are
|
||||
|
||||
trying to compensate for the insulin
|
||||
|
||||
resistance this means that the insulin
|
||||
|
||||
glucagon balances such that diabetic
|
||||
|
||||
ketoacidosis does not usually develop
|
||||
|
||||
having said that a complication called
|
||||
|
||||
hyperosmolar hyperglycemic state or HHS
|
||||
|
||||
is much more common in type 2 diabetes
|
||||
|
||||
than type 1 diabetes and it causes
|
||||
|
||||
increased plasma osmolarity due to
|
||||
|
||||
extreme dehydration and concentration of
|
||||
|
||||
the blood to help understand this
|
||||
|
||||
remember that glucose is a polar
|
||||
|
||||
molecule that cannot passively diffuse
|
||||
|
||||
across cell membranes which means that
|
||||
|
||||
it acts as a solute so when levels of
|
||||
|
||||
glucose are super high in the blood
|
||||
|
||||
meaning it's a hyperosmolar State water
|
||||
|
||||
starts to leave the body cells and enter
|
||||
|
||||
the blood vessels leaving the cells were
|
||||
|
||||
relatively dry in travailed rather than
|
||||
|
||||
plump and juicy blood vessels that are
|
||||
|
||||
full of water lead to increased
|
||||
|
||||
urination and total body dehydration and
|
||||
|
||||
this is a very serious situation because
|
||||
|
||||
the dehydration of the body's cells and
|
||||
|
||||
in particular the brain can cause a
|
||||
|
||||
number of symptoms including mental
|
||||
|
||||
status changes in HHS you can sometimes
|
||||
|
||||
see mild ketone emia and acidosis but
|
||||
|
||||
not to the extent that it's seen in DKA
|
||||
|
||||
and in DKA you can see some hyper
|
||||
|
||||
osmolarity so there's definitely overlap
|
||||
|
||||
between these two syndromes
|
||||
|
||||
besides type 1 and type 2 diabetes there
|
||||
|
||||
are also a couple other subtypes of
|
||||
|
||||
diabetes mellitus gestational diabetes
|
||||
|
||||
is when pregnant women have increased
|
||||
|
||||
blood glucose which is particularly
|
||||
|
||||
during the third trimester although
|
||||
|
||||
ultimately unknown the cause is thought
|
||||
|
||||
to be related to pregnancy hormones that
|
||||
|
||||
interfere with insulins action on
|
||||
|
||||
insulin receptors also sometimes people
|
||||
|
||||
can develop drug-induced diabetes which
|
||||
|
||||
is where medications have side effects
|
||||
|
||||
that tend to increase blood glucose
|
||||
|
||||
levels the mechanism for both of these
|
||||
|
||||
is thought to be related to insulin
|
||||
|
||||
resistance like type 2 diabetes rather
|
||||
|
||||
than an autoimmune destruction process
|
||||
|
||||
like in type 1 diabetes diagnosing type
|
||||
|
||||
1 or type 2 diabetes is done by getting
|
||||
|
||||
a sense for how much glucose is floating
|
||||
|
||||
around in the blood and has specific
|
||||
|
||||
standards that the World Health
|
||||
|
||||
Organization uses very commonly a
|
||||
|
||||
fasting glucose test is taken where the
|
||||
|
||||
person doesn't eat or drink except the
|
||||
|
||||
water that's okay for a total of eight
|
||||
|
||||
hours and then has their blood tested
|
||||
|
||||
for glucose levels levels of 100
|
||||
|
||||
milligrams per deciliter to 120
|
||||
|
||||
five milligrams per deciliter indicates
|
||||
|
||||
pre-diabetes and 126 milligrams per
|
||||
|
||||
deciliter or higher indicates diabetes a
|
||||
|
||||
non fasting a random glucose test can be
|
||||
|
||||
done at any time with 200 milligrams per
|
||||
|
||||
deciliter or higher being a red flag for
|
||||
|
||||
diabetes another test is called an oral
|
||||
|
||||
glucose tolerance test where person is
|
||||
|
||||
given glucose and then blood samples are
|
||||
|
||||
taken at time intervals to figure out
|
||||
|
||||
how well it's being cleared from the
|
||||
|
||||
blood the most important interval being
|
||||
|
||||
two hours later levels of 140 milligrams
|
||||
|
||||
per deciliter to 199 milligrams per
|
||||
|
||||
deciliter indicate pre-diabetes
|
||||
|
||||
and 200 or above indicates diabetes
|
||||
|
||||
another thing to know is that when blood
|
||||
|
||||
glucose levels get high the glucose can
|
||||
|
||||
also stick to proteins that are floating
|
||||
|
||||
around in the blood or in cells so that
|
||||
|
||||
brings us to another type of test that
|
||||
|
||||
can be done which is the hba1c test
|
||||
|
||||
which tests for the proportion of
|
||||
|
||||
hemoglobin in red blood cells that has
|
||||
|
||||
glucose stuck to it called glycated
|
||||
|
||||
hemoglobin hba1c levels of 5.7% 26.4%
|
||||
|
||||
indicate pre-diabetes
|
||||
|
||||
and 6.5 percent or higher indicates
|
||||
|
||||
diabetes this proportion of glycated
|
||||
|
||||
hemoglobin doesn't change day to day so
|
||||
|
||||
it gives a sense for whether the blood
|
||||
|
||||
glucose levels have been high over the
|
||||
|
||||
past two to three months finally we have
|
||||
|
||||
the c-peptide test which tests for
|
||||
|
||||
byproducts of insulin production if the
|
||||
|
||||
level of c-peptide is low or absent it
|
||||
|
||||
means the pancreas is no longer
|
||||
|
||||
producing enough insulin and the glucose
|
||||
|
||||
cannot enter the cells
|
||||
|
||||
for type one diabetes insulin is the
|
||||
|
||||
only treatment option for type 2
|
||||
|
||||
diabetes on the other hand lifestyle
|
||||
|
||||
changes like weight loss and exercise
|
||||
|
||||
along with a healthy diet and an oral
|
||||
|
||||
anti-diabetic medication like metformin
|
||||
|
||||
in several other classes can sometimes
|
||||
|
||||
be enough to reverse some of that
|
||||
|
||||
insulin resistance and keep blood sugar
|
||||
|
||||
levels in check however if oral
|
||||
|
||||
anti-diabetic medications fail type 2
|
||||
|
||||
diabetes can also be treated with
|
||||
|
||||
insulin something to bear in mind is
|
||||
|
||||
that insulin treatment comes with a risk
|
||||
|
||||
of hypoglycemia especially if insulin is
|
||||
|
||||
taken without a meal symptoms of
|
||||
|
||||
hypoglycemia can be mild like weakness
|
||||
|
||||
hunger and shaking but they can progress
|
||||
|
||||
to a loss of consciousness in seizures
|
||||
|
||||
in severe cases in mild cases drinking
|
||||
|
||||
juices or eating candy or sugar might be
|
||||
|
||||
enough to bring blood sugar up but in
|
||||
|
||||
severe cases intravenous glucose should
|
||||
|
||||
be given as soon as possible
|
||||
|
||||
the FDA has also recently approved
|
||||
|
||||
intranasal glucagon as a treatment for
|
||||
|
||||
severe hypoglycemia all right now over
|
||||
|
||||
time high glucose levels can cause
|
||||
|
||||
damage to tiny blood vessels while the
|
||||
|
||||
micro vasculature in arterioles a
|
||||
|
||||
process called hyaline
|
||||
|
||||
arteriolosclerosis is where the walls of
|
||||
|
||||
the arterioles develop hyaline deposits
|
||||
|
||||
which are deposits of proteins and these
|
||||
|
||||
make them hard and inflexible in
|
||||
|
||||
capillaries the basement membrane can
|
||||
|
||||
thicken and make it difficult for oxygen
|
||||
|
||||
to easily move from the capillary to the
|
||||
|
||||
tissues causing hypoxia
|
||||
|
||||
one of the most significant effects is
|
||||
|
||||
that diabetes increases the risk of
|
||||
|
||||
medium and large arterial wall damage
|
||||
|
||||
and subsequent atherosclerosis which can
|
||||
|
||||
lead to heart attacks and strokes which
|
||||
|
||||
are major causes of morbidity and
|
||||
|
||||
mortality for patients with diabetes in
|
||||
|
||||
the eyes diabetes can lead to
|
||||
|
||||
retinopathy and evidence of that can be
|
||||
|
||||
seen on a fundus copic exam that shows
|
||||
|
||||
cotton-wool spots or flare hemorrhages
|
||||
|
||||
and can eventually cause blindness in
|
||||
|
||||
the kidneys the a ferrant and efferent
|
||||
|
||||
arterioles as well as the glomerulus
|
||||
|
||||
itself can get damaged which can lead to
|
||||
|
||||
an F Radek syndrome that slowly
|
||||
|
||||
diminishes the kidneys ability to filter
|
||||
|
||||
blood over time and can ultimately lead
|
||||
|
||||
to dialysis diabetes can also affect the
|
||||
|
||||
function of nerves causing symptoms like
|
||||
|
||||
a decrease in sensation in the toes and
|
||||
|
||||
fingers sometimes called a stocking
|
||||
|
||||
glove distribution as well as causes the
|
||||
|
||||
autonomic nervous system to malfunction
|
||||
|
||||
and that system controls a number of
|
||||
|
||||
body functions
|
||||
|
||||
everything from sweating to passing gas
|
||||
|
||||
finally both the poor blood supply and
|
||||
|
||||
nerve damage can lead to ulcers
|
||||
|
||||
typically on the feet that don't heal
|
||||
|
||||
quickly and can get pretty severe and
|
||||
|
||||
need to be amputated these are some of
|
||||
|
||||
the complications of uncontrolled
|
||||
|
||||
diabetes which is why it's important to
|
||||
|
||||
diagnose and control diabetes through a
|
||||
|
||||
healthy lifestyle medications to reduce
|
||||
|
||||
insulin resistance and even insulin
|
||||
|
||||
therapy if beta cells have been
|
||||
|
||||
exhausted while type 1 diabetes cannot
|
||||
|
||||
be prevented type 2 diabetes can in fact
|
||||
|
||||
many people with diabetes can control
|
||||
|
||||
their blood sugar levels really
|
||||
|
||||
effectively and live a full and active
|
||||
|
||||
life without any of the complications
|
||||
|
||||
thanks for watching if you're interested
|
||||
|
||||
in a deeper dive on this topic take a
|
||||
|
||||
look at as Moses org where we have
|
||||
|
||||
flashcards questions and other awesome
|
||||
|
||||
tools to help you learn medicine
|
||||
|
||||
you
|
||||
|
||||
File diff suppressed because one or more lines are too long
File diff suppressed because it is too large
Load Diff
@@ -1,34 +0,0 @@
|
||||
import os
|
||||
import subprocess
|
||||
import sys
|
||||
|
||||
from loguru import logger
|
||||
|
||||
# Get the input file name from the command line argument
|
||||
input_file = sys.argv[1]
|
||||
# example use: python 0-reflector-local.py input.m4a agenda.txt
|
||||
|
||||
# Get the agenda file name from the command line argument if provided
|
||||
if len(sys.argv) > 2:
|
||||
agenda_file = sys.argv[2]
|
||||
else:
|
||||
agenda_file = "agenda.txt"
|
||||
# example use: python 0-reflector-local.py input.m4a my_agenda.txt
|
||||
|
||||
# Check if the agenda file exists
|
||||
if not os.path.exists(agenda_file):
|
||||
logger.error("agenda_file is missing")
|
||||
|
||||
# Check if the input file is .m4a, if so convert to .mp4
|
||||
if input_file.endswith(".m4a"):
|
||||
subprocess.run(["ffmpeg", "-i", input_file, f"{input_file}.mp4"])
|
||||
input_file = f"{input_file}.mp4"
|
||||
|
||||
# Run the first script to generate the transcript
|
||||
subprocess.run(["python3", "1-transcript-generator.py", input_file, f"{input_file}_transcript.txt"])
|
||||
|
||||
# Run the second script to compare the transcript to the agenda
|
||||
subprocess.run(["python3", "2-agenda-transcript-diff.py", agenda_file, f"{input_file}_transcript.txt"])
|
||||
|
||||
# Run the third script to summarize the transcript
|
||||
subprocess.run(["python3", "3-transcript-summarizer.py", f"{input_file}_transcript.txt", f"{input_file}_summary.txt"])
|
||||
@@ -1,62 +0,0 @@
|
||||
import argparse
|
||||
import os
|
||||
|
||||
import moviepy.editor
|
||||
import whisper
|
||||
from loguru import logger
|
||||
|
||||
WHISPER_MODEL_SIZE = "base"
|
||||
|
||||
|
||||
def init_argparse() -> argparse.ArgumentParser:
|
||||
parser = argparse.ArgumentParser(
|
||||
usage="%(prog)s <LOCATION> <OUTPUT>",
|
||||
description="Creates a transcript of a video or audio file using the OpenAI Whisper model"
|
||||
)
|
||||
parser.add_argument("location", help="Location of the media file")
|
||||
parser.add_argument("output", help="Output file path")
|
||||
return parser
|
||||
|
||||
|
||||
def main():
|
||||
import sys
|
||||
sys.setrecursionlimit(10000)
|
||||
|
||||
parser = init_argparse()
|
||||
args = parser.parse_args()
|
||||
|
||||
media_file = args.location
|
||||
logger.info(f"Processing file: {media_file}")
|
||||
|
||||
# Check if the media file is a valid audio or video file
|
||||
if os.path.isfile(media_file) and not media_file.endswith(
|
||||
('.mp3', '.wav', '.ogg', '.flac', '.mp4', '.avi', '.flv')):
|
||||
logger.error(f"Invalid file format: {media_file}")
|
||||
return
|
||||
|
||||
# If the media file we just retrieved is an audio file then skip extraction step
|
||||
audio_filename = media_file
|
||||
logger.info(f"Found audio-only file, skipping audio extraction")
|
||||
|
||||
audio = moviepy.editor.AudioFileClip(audio_filename)
|
||||
|
||||
logger.info("Selected extracted audio")
|
||||
|
||||
# Transcribe the audio file using the OpenAI Whisper model
|
||||
logger.info("Loading Whisper speech-to-text model")
|
||||
whisper_model = whisper.load_model(WHISPER_MODEL_SIZE)
|
||||
|
||||
logger.info(f"Transcribing file: {media_file}")
|
||||
whisper_result = whisper_model.transcribe(media_file)
|
||||
|
||||
logger.info("Finished transcribing file")
|
||||
|
||||
# Save the transcript to the specified file.
|
||||
logger.info(f"Saving transcript to: {args.output}")
|
||||
transcript_file = open(args.output, "w")
|
||||
transcript_file.write(whisper_result["text"])
|
||||
transcript_file.close()
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,68 +0,0 @@
|
||||
import argparse
|
||||
|
||||
import spacy
|
||||
from loguru import logger
|
||||
|
||||
|
||||
# Define the paths for agenda and transcription files
|
||||
def init_argparse() -> argparse.ArgumentParser:
|
||||
parser = argparse.ArgumentParser(
|
||||
usage="%(prog)s <AGENDA> <TRANSCRIPTION>",
|
||||
description="Compares the transcript of a video or audio file to an agenda using the SpaCy model"
|
||||
)
|
||||
parser.add_argument("agenda", help="Location of the agenda file")
|
||||
parser.add_argument("transcription", help="Location of the transcription file")
|
||||
return parser
|
||||
|
||||
|
||||
args = init_argparse().parse_args()
|
||||
agenda_path = args.agenda
|
||||
transcription_path = args.transcription
|
||||
|
||||
# Load the spaCy model and add the sentencizer
|
||||
spaCy_model = "en_core_web_md"
|
||||
nlp = spacy.load(spaCy_model)
|
||||
nlp.add_pipe('sentencizer')
|
||||
logger.info("Loaded spaCy model " + spaCy_model)
|
||||
|
||||
# Load the agenda
|
||||
with open(agenda_path, "r") as f:
|
||||
agenda = [line.strip() for line in f.readlines() if line.strip()]
|
||||
logger.info("Loaded agenda items")
|
||||
|
||||
# Load the transcription
|
||||
with open(transcription_path, "r") as f:
|
||||
transcription = f.read()
|
||||
logger.info("Loaded transcription")
|
||||
|
||||
# Tokenize the transcription using spaCy
|
||||
doc_transcription = nlp(transcription)
|
||||
logger.info("Tokenized transcription")
|
||||
|
||||
# Find the items covered in the transcription
|
||||
covered_items = {}
|
||||
for item in agenda:
|
||||
item_doc = nlp(item)
|
||||
for sent in doc_transcription.sents:
|
||||
if not sent or not all(token.has_vector for token in sent):
|
||||
# Skip an empty span or one without any word vectors
|
||||
continue
|
||||
similarity = sent.similarity(item_doc)
|
||||
similarity_threshold = 0.7
|
||||
if similarity > similarity_threshold: # Set the threshold to determine what is considered a match
|
||||
covered_items[item] = True
|
||||
break
|
||||
|
||||
# Count the number of items covered and calculatre the percentage
|
||||
num_covered_items = sum(covered_items.values())
|
||||
percentage_covered = num_covered_items / len(agenda) * 100
|
||||
|
||||
# Print the results
|
||||
print("💬 Agenda items covered in the transcription:")
|
||||
for item in agenda:
|
||||
if item in covered_items and covered_items[item]:
|
||||
print("✅ ", item)
|
||||
else:
|
||||
print("❌ ", item)
|
||||
print("📊 Coverage: {:.2f}%".format(percentage_covered))
|
||||
logger.info("Finished comparing agenda to transcription with similarity threshold of " + str(similarity_threshold))
|
||||
@@ -1,94 +0,0 @@
|
||||
import argparse
|
||||
|
||||
import nltk
|
||||
|
||||
nltk.download('stopwords')
|
||||
from nltk.corpus import stopwords
|
||||
from nltk.tokenize import word_tokenize, sent_tokenize
|
||||
from heapq import nlargest
|
||||
from loguru import logger
|
||||
|
||||
|
||||
# Function to initialize the argument parser
|
||||
def init_argparse():
|
||||
parser = argparse.ArgumentParser(
|
||||
usage="%(prog)s <TRANSCRIPT> <SUMMARY>",
|
||||
description="Summarization"
|
||||
)
|
||||
parser.add_argument("transcript", type=str, default="transcript.txt", help="Path to the input transcript file")
|
||||
parser.add_argument("summary", type=str, default="summary.txt", help="Path to the output summary file")
|
||||
parser.add_argument("--num_sentences", type=int, default=5, help="Number of sentences to include in the summary")
|
||||
return parser
|
||||
|
||||
|
||||
# Function to read the input transcript file
|
||||
def read_transcript(file_path):
|
||||
with open(file_path, "r") as file:
|
||||
transcript = file.read()
|
||||
return transcript
|
||||
|
||||
|
||||
# Function to preprocess the text by removing stop words and special characters
|
||||
def preprocess_text(text):
|
||||
stop_words = set(stopwords.words('english'))
|
||||
words = word_tokenize(text)
|
||||
words = [w.lower() for w in words if w.isalpha() and w.lower() not in stop_words]
|
||||
return words
|
||||
|
||||
|
||||
# Function to score each sentence based on the frequency of its words and return the top sentences
|
||||
def summarize_text(text, num_sentences):
|
||||
# Tokenize the text into sentences
|
||||
sentences = sent_tokenize(text)
|
||||
|
||||
# Preprocess the text by removing stop words and special characters
|
||||
words = preprocess_text(text)
|
||||
|
||||
# Calculate the frequency of each word in the text
|
||||
word_freq = nltk.FreqDist(words)
|
||||
|
||||
# Calculate the score for each sentence based on the frequency of its words
|
||||
sentence_scores = {}
|
||||
for i, sentence in enumerate(sentences):
|
||||
sentence_words = preprocess_text(sentence)
|
||||
for word in sentence_words:
|
||||
if word in word_freq:
|
||||
if i not in sentence_scores:
|
||||
sentence_scores[i] = word_freq[word]
|
||||
else:
|
||||
sentence_scores[i] += word_freq[word]
|
||||
|
||||
# Select the top sentences based on their scores
|
||||
top_sentences = nlargest(num_sentences, sentence_scores, key=sentence_scores.get)
|
||||
|
||||
# Sort the top sentences in the order they appeared in the original text
|
||||
summary_sent = sorted(top_sentences)
|
||||
summary = [sentences[i] for i in summary_sent]
|
||||
|
||||
return " ".join(summary)
|
||||
|
||||
|
||||
def main():
|
||||
# Initialize the argument parser and parse the arguments
|
||||
parser = init_argparse()
|
||||
args = parser.parse_args()
|
||||
|
||||
# Read the input transcript file
|
||||
logger.info(f"Reading transcript from: {args.transcript}")
|
||||
transcript = read_transcript(args.transcript)
|
||||
|
||||
# Summarize the transcript using the nltk library
|
||||
logger.info("Summarizing transcript")
|
||||
summary = summarize_text(transcript, args.num_sentences)
|
||||
|
||||
# Write the summary to the output file
|
||||
logger.info(f"Writing summary to: {args.summary}")
|
||||
with open(args.summary, "w") as f:
|
||||
f.write("Summary of: " + args.transcript + "\n\n")
|
||||
f.write(summary)
|
||||
|
||||
logger.info("Summarization completed")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,4 +0,0 @@
|
||||
# Deloitte HR @ NYS Cybersecurity Conference
|
||||
- ways to retain and grow your workforce
|
||||
- how to enable cybersecurity professionals to do their best work
|
||||
- low-budget activities that can be implemented starting tomorrow
|
||||
File diff suppressed because one or more lines are too long
@@ -1,3 +0,0 @@
|
||||
Summary of: 30min-CyberHR/30min-CyberHR.m4a.mp4_transcript.txt
|
||||
|
||||
Since the workforce is an organization's most valuable asset, investing in workforce experience activities, we've found has lead to more productive work, more efficient work, more innovative approaches to the work, and more engaged teams which ultimately results in better mission outcomes for your organization. And this one really focuses on not just pulsing a workforce once a year through an annual HR survey of, how do you really feel like, you know, what leadership considerations should we implement or, you know, how can we enhance the performance management process. We've just found that, you know, by investing in this and putting the workforce as, you know, the center part of what you invest in as an organization and leaders, it's not only about retention, talent, you know, the cyber workforce crisis, but people want to do work well and they're able to get more done and achieve more without you, you know, directly supervising and micromanaging or looking at everything because, you know, you know, you know, you're not going to be able to do anything. I hope there was a little bit of, you know, the landscape of the cyber workforce with some practical tips that you can take away for how to just think about, you know, improving the overall workforce experience and investing in your employees. So with this, you know, we know that all of you are in the trenches every day, you're facing this, you're living this, and we are just interested to hear from all of you, you know, just to start, like, what's one thing that has worked well in your organization in terms of enhancing or investing in the workforce experience?
|
||||
File diff suppressed because one or more lines are too long
@@ -1,47 +0,0 @@
|
||||
AGENDA: Most important things to look for in a start up
|
||||
|
||||
TAM: Make sure the market is sufficiently large than once they win they can get rewarded
|
||||
- Medium sized markets that should be winner take all can work
|
||||
- TAM needs to be realistic of direct market size
|
||||
|
||||
Product market fit: Being in a good market with a product than can satisfy that market
|
||||
- Solves a problem
|
||||
- Builds a solution a customer wants to buy
|
||||
- Either saves the customer something (time/money/pain) or gives them something (revenue/enjoyment)
|
||||
|
||||
Unit economics: Profit for delivering all-in cost must be attractive (% or $ amount)
|
||||
- Revenue minus direct costs
|
||||
- Raw input costs (materials, variable labour), direct cost of delivering and servicing the sale
|
||||
- Attractive as a % of sales so it can contribute to fixed overhead
|
||||
- Look for high incremental contribution margin
|
||||
|
||||
LTV CAC: Life-time value (revenue contribution) vs cost to acquire customer must be healthy
|
||||
- LTV = Purchase value x number of purchases x customer lifespan
|
||||
- CAC = All-in costs of sales + marketing over number of new customer additions
|
||||
- Strong reputation leads to referrals leads to lower CAC. Want customers evangelizing product/service
|
||||
- Rule of thumb higher than 3
|
||||
|
||||
Churn: Fits into LTV, low churn leads to higher LTV and helps keep future CAC down
|
||||
- Selling to replenish revenue every year is hard
|
||||
- Can run through entire customer base over time
|
||||
- Low churn builds strong net dollar retention
|
||||
|
||||
Business: Must have sufficient barriers to entry to ward off copy-cats once established
|
||||
- High switching costs (lock-in)
|
||||
- Addictive
|
||||
- Steep learning curve once adopted (form of switching cost)
|
||||
- Two sided liquidity
|
||||
- Patents, IP, Branding
|
||||
- No hyper-scaler who can roll over you quickly
|
||||
- Scale could be a barrier to entry but works against most start-ups, not for them
|
||||
- Once developed, answer question: Could a well funded competitor starting up today easily duplicate this business or is it cheaper to buy the start up?
|
||||
|
||||
Founders: Must be religious about their product. Believe they will change the world against all odds.
|
||||
- Just money in the bank is not enough to build a successful company. Just good tech not enough
|
||||
to build a successful company
|
||||
- Founders must be motivated to build something, not (all) about money. They would be doing
|
||||
this for free because they believe in it. Not looking for quick score
|
||||
- Founders must be persuasive. They will be asking others to sacrifice to make their dream come
|
||||
to life. They will need to convince investors this company can work and deserves funding.
|
||||
- Must understand who the customer is and what problem they are helping to solve.
|
||||
- Founders aren’t expected to know all the preceding points in this document but have an understanding of most of this, and be able to offer a vision.
|
||||
@@ -1,8 +0,0 @@
|
||||
AGENDA: Most important things to look for in a start up
|
||||
TAM: Make sure the market is sufficiently large than once they win they can get rewarded
|
||||
Product market fit: Being in a good market with a product than can satisfy that market
|
||||
Unit economics: Profit for delivering all-in cost must be attractive (% or $ amount)
|
||||
LTV CAC: Life-time value (revenue contribution) vs cost to acquire customer must be healthy
|
||||
Churn: Fits into LTV, low churn leads to higher LTV and helps keep future CAC down
|
||||
Business: Must have sufficient barriers to entry to ward off copy-cats once established
|
||||
Founders: Must be religious about their product. Believe they will change the world against all odds.
|
||||
@@ -1,10 +0,0 @@
|
||||
Summary of: recordings/42min-StartupsTechTalk.mp4
|
||||
|
||||
The speaker discusses their plan to launch an investment company, which will sit on a pool of cash raised from various partners and investors. They will take equity stakes in startups that they believe have the potential to scale and become successful. The speaker emphasizes the importance of investing in companies that have a large total addressable market (TAM) and good product-market fit. They also discuss the concept of unit economics and how it is important to ensure that the profit from selling a product or service outweighs the cost of producing it. The speaker encourages their team to keep an eye out for interesting startups and to send them their way if they come across any.
|
||||
|
||||
The conversation is about the importance of unit economics, incremental margin, lifetime value, customer acquisition costs, churn, and barriers to entry in evaluating businesses for investment. The speaker explains that companies with good unit economics and high incremental contribution margins are ideal for investment. Lifetime value measures how much a customer will spend on a business over their entire existence, while customer acquisition costs measure the cost of acquiring a new customer. Churn refers to the rate at which customers leave a business, and businesses with low churn tend to have high lifetime values. High barriers to entry, such as high switching costs, can make it difficult for competitors to enter the market and kill established businesses.
|
||||
|
||||
The speaker discusses various factors that can contribute to a company's success and create a competitive advantage. These include making the product addictive, having steep learning curves, creating two-sided liquidity for marketplaces, having patents or intellectual property, strong branding, and scale as a barrier to entry. The speaker also emphasizes the importance of founders having a plan to differentiate themselves from competitors and avoid being rolled over by larger companies. Additionally, the speaker mentions MasterCard and Visa as examples of companies that invented their markets, while Apple was able to build a strong brand despite starting with no developers or users.
|
||||
|
||||
The speaker discusses the importance of founders in building successful companies, emphasizing that they must be passionate and believe in their product. They should also be charismatic and able to persuade others to work towards their vision. The speaker cites examples of successful CEOs such as Zuckerberg, Steve Jobs, Elon Musk, Bill Gates, Jeff Bezos, Travis Kalanick, and emphasizes that luck is also a factor in success. The speaker encourages listeners to have a critical eye when evaluating startups and to look for those with a clear understanding of their customers and the problem they are solving.
|
||||
|
||||
File diff suppressed because one or more lines are too long
@@ -1,3 +0,0 @@
|
||||
Summary of: 42min-StartupsTechTalk/42min-StartupsTechTalk.mp4_transcript.txt
|
||||
|
||||
If you had perfect knowledge, and you need like one more piece of advertising, drove like 0.2 customers in each customer generates, like let's say you wanted to completely maximize, you'd make it say your contribution margin, on incremental sales, is just over what you're spending on ad revenue. Like if you're, I don't know, well, let's see, I got like you don't really want to advertise a ton in the huge and everywhere, and then getting to ubiquitous, because you grab it, damage your brands, but just like an economic textbook theory, and be like, it'd be that basic math. And the table's like exactly, we're going to be really cautious to like be able to move in a year if we need to, but Google's goal is going to be giving away foundational models, lock everyone in, make them use Google Cloud, make them use Google Tools, and it's going to be very hard to switch off. Like if you were starting to develop Figma, you might say, okay, well Adobe is just gonna eat my lunch, right, like right away. So when you see a startup or talk to a founder and he's saying these things in your head like, man, this isn't gonna work because of, you know, there's no tab or there's, you know, like Amazon's gonna roll these cuts over in like two days or whatever, you know, or the man, this is really interesting because not only they're not doing it and no one else is doing this, but like they're going after a big market.
|
||||
File diff suppressed because one or more lines are too long
@@ -1,4 +0,0 @@
|
||||
GitHub
|
||||
Requirements
|
||||
Junior Developers
|
||||
Riding Elephants
|
||||
@@ -1,4 +0,0 @@
|
||||
Summary of: https://www.youtube.com/watch?v=DzRoYc2UGKI
|
||||
|
||||
Small Developer is a program that creates an entire project for you based on a prompt. It uses the JATGPT API to generate code and files, and it's easy to use. The program can be installed by cloning the GitHub repository and using modalcom. The program can create projects for various languages, including Python and Ruby. You can also create a prompt.md file to input your prompt instead of pasting it into the terminal. The program is useful for creating detailed specs that can be passed on to junior developers. Overall, Small Developer is a helpful tool for quickly generating code and projects.
|
||||
|
||||
File diff suppressed because one or more lines are too long
@@ -1,11 +0,0 @@
|
||||
# Record on Voice Memos on iPhone
|
||||
|
||||
# Airdrop to MacBook Air
|
||||
|
||||
# Run Reflector on .m4a Recording and Agenda
|
||||
|
||||
python 0-reflector-local.py voicememo.m4a agenda.txt
|
||||
|
||||
OR - using 30min-CyberHR example:
|
||||
|
||||
python 0-reflector-local.py 30min-CyberHR/30min-CyberHR.m4a 30min-CyberHR/30min-CyberHR-agenda.txt
|
||||
@@ -1,125 +0,0 @@
|
||||
import argparse
|
||||
import os
|
||||
import tempfile
|
||||
|
||||
import moviepy.editor
|
||||
import nltk
|
||||
import whisper
|
||||
from loguru import logger
|
||||
from transformers import BartTokenizer, BartForConditionalGeneration
|
||||
|
||||
nltk.download('punkt', quiet=True)
|
||||
|
||||
WHISPER_MODEL_SIZE = "base"
|
||||
|
||||
|
||||
def init_argparse() -> argparse.ArgumentParser:
|
||||
parser = argparse.ArgumentParser(
|
||||
usage="%(prog)s [OPTIONS] <LOCATION> <OUTPUT>",
|
||||
description="Creates a transcript of a video or audio file, then summarizes it using BART."
|
||||
)
|
||||
|
||||
parser.add_argument("location", help="Location of the media file")
|
||||
parser.add_argument("output", help="Output file path")
|
||||
|
||||
parser.add_argument(
|
||||
"-t", "--transcript", help="Save a copy of the intermediary transcript file", type=str)
|
||||
parser.add_argument(
|
||||
"-l", "--language", help="Language that the summary should be written in",
|
||||
type=str, default="english", choices=['english', 'spanish', 'french', 'german', 'romanian'])
|
||||
parser.add_argument(
|
||||
"-m", "--model_name", help="Name or path of the BART model",
|
||||
type=str, default="facebook/bart-large-cnn")
|
||||
|
||||
return parser
|
||||
|
||||
|
||||
# NLTK chunking function
|
||||
def chunk_text(txt, max_chunk_length=500):
|
||||
"Split text into smaller chunks."
|
||||
sentences = nltk.sent_tokenize(txt)
|
||||
chunks = []
|
||||
current_chunk = ""
|
||||
for sentence in sentences:
|
||||
if len(current_chunk) + len(sentence) < max_chunk_length:
|
||||
current_chunk += f" {sentence.strip()}"
|
||||
else:
|
||||
chunks.append(current_chunk.strip())
|
||||
current_chunk = f"{sentence.strip()}"
|
||||
chunks.append(current_chunk.strip())
|
||||
return chunks
|
||||
|
||||
|
||||
# BART summary function
|
||||
def summarize_chunks(chunks, tokenizer, model):
|
||||
summaries = []
|
||||
for c in chunks:
|
||||
input_ids = tokenizer.encode(c, return_tensors='pt')
|
||||
summary_ids = model.generate(
|
||||
input_ids, num_beams=4, length_penalty=2.0, max_length=1024, no_repeat_ngram_size=3)
|
||||
summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)
|
||||
summaries.append(summary)
|
||||
return summaries
|
||||
|
||||
|
||||
def main():
|
||||
import sys
|
||||
sys.setrecursionlimit(10000)
|
||||
|
||||
parser = init_argparse()
|
||||
args = parser.parse_args()
|
||||
|
||||
media_file = args.location
|
||||
logger.info(f"Processing file: {media_file}")
|
||||
|
||||
# If the media file we just retrieved is a video, extract its audio stream.
|
||||
if os.path.isfile(media_file) and media_file.endswith(('.mp4', '.avi', '.flv')):
|
||||
audio_filename = tempfile.NamedTemporaryFile(
|
||||
suffix=".mp3", delete=False).name
|
||||
logger.info(f"Extracting audio to: {audio_filename}")
|
||||
|
||||
video = moviepy.editor.VideoFileClip(media_file)
|
||||
video.audio.write_audiofile(audio_filename, logger=None)
|
||||
|
||||
logger.info("Finished extracting audio")
|
||||
media_file = audio_filename
|
||||
|
||||
# Transcribe the audio file using the OpenAI Whisper model
|
||||
logger.info("Loading Whisper speech-to-text model")
|
||||
whisper_model = whisper.load_model(WHISPER_MODEL_SIZE)
|
||||
|
||||
logger.info(f"Transcribing audio file: {media_file}")
|
||||
whisper_result = whisper_model.transcribe(media_file)
|
||||
|
||||
logger.info("Finished transcribing file")
|
||||
|
||||
# If we got the transcript parameter on the command line, save the transcript to the specified file.
|
||||
if args.transcript:
|
||||
logger.info(f"Saving transcript to: {args.transcript}")
|
||||
transcript_file = open(args.transcript, "w")
|
||||
transcript_file.write(whisper_result["text"])
|
||||
transcript_file.close()
|
||||
|
||||
# Summarize the generated transcript using the BART model
|
||||
logger.info(f"Loading BART model: {args.model_name}")
|
||||
tokenizer = BartTokenizer.from_pretrained(args.model_name)
|
||||
model = BartForConditionalGeneration.from_pretrained(args.model_name)
|
||||
|
||||
logger.info("Breaking transcript into smaller chunks")
|
||||
chunks = chunk_text(whisper_result['text'])
|
||||
|
||||
logger.info(
|
||||
f"Transcript broken into {len(chunks)} chunks of at most 500 words") # TODO fix variable
|
||||
|
||||
logger.info(f"Writing summary text in {args.language} to: {args.output}")
|
||||
with open(args.output, 'w') as f:
|
||||
f.write('Summary of: ' + args.location + "\n\n")
|
||||
summaries = summarize_chunks(chunks, tokenizer, model)
|
||||
for summary in summaries:
|
||||
f.write(summary.strip() + "\n\n")
|
||||
|
||||
logger.info("Summarization completed")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,24 +0,0 @@
|
||||
# Steps to prepare data and submit/check OpenAI finetuning
|
||||
# import subprocess
|
||||
# subprocess.run("openai tools fine_tunes.prepare_data -f " + "finetuning_dataset.jsonl")
|
||||
# export OPENAI_API_KEY=
|
||||
# openai api fine_tunes.create -t <TRAIN_FILE_ID_OR_PATH> -m <BASE_MODEL>
|
||||
# openai api fine_tunes.list
|
||||
|
||||
|
||||
import openai
|
||||
|
||||
# Use your OpenAI API Key
|
||||
openai.api_key = ""
|
||||
|
||||
sample_chunks = ["You all just came off of your incredible Google Cloud next conference where you released a wide variety of functionality and features and new products across artisan television and also across the entire sort of cloud ecosystem . You want to just first by walking through , first start by walking through all the innovations that you sort of released and what you 're excited about when you come to Google Cloud ? Now our vision is super simple . If you look at what smartphones did for a consumer , you know they took a computer and internet browser , a communication device , and a camera , and made it so that it 's in everybody 's pocket , so it really brought computation to every person . We feel that , you know , our , what we 're trying to do is take all the technological innovation that Google 's doing , but make it super simple so that everyone can consume it . And so that includes our global data center footprint , all the new types of hardware and large-scale systems we work on , the software that we 're making available for people to do high-scale computation , tools for data processing , tools for cybersecurity , processing , tools for cyber security , tools for machine learning , but make it so simple that everyone can use it . And every step that we do to simplify things for people , we think adoption can grow . And so that 's a lot of what we 've done these last three , four years , and we made a number of announcements that next in machine learning and AI in particular , you know , we look at our work as four elements , how we take our large-scale compute systems that were building for AI and how we make that available to everybody . Second , what we 're doing with the software stacks and top of it , things like jacks and other things and how we 're making those available to everybody . Third is advances because different people have different levels of expertise . Some people say I need the hardware to build my own large language model or algorithm . Other people say , look , I really need to use a building block . You guys give me . So , 30s we 've done a lot with AutoML and we announce new capability for image , video , and translation to make it available to everybody . And then lastly , we 're also building completely packaged solutions for some areas and we announce some new stuff . -> ",
|
||||
" We 're joined next by Thomas Curian , CEO of Google Cloud , and Alexander Wang , CEO and founder of Scale AI . Thomas joined Google in November 2018 as the CEO of Google Cloud . Prior to Google , Thomas spent 22 years at Oracle , where most recently he was president of product development . Before that , Thomas worked at McKinsey as a business analyst and engagement manager . His nearly 30 years of experience have given him a deep knowledge of engineering enterprise relationships and leadership of large organizations . Thomas 's degrees include an MBA in administration and management from Stanford University , as an RJ Miller scholar and a BSEE in electrical engineering and computer science from Princeton University , where he graduated suma cum laude . Thomas serves as a member of the Stanford graduate School of Business Advisory Council and Princeton University School of Engineering Advisory Council . Please welcome to the stage , Thomas Curian and Alexander Wang . This is a super exciting conversation . Thanks for being here , Thomas . - > "]
|
||||
|
||||
# Give your finetuned model name here
|
||||
# "davinci:ft-personal-2023-07-14-10-43-51"
|
||||
model_name = ""
|
||||
response = openai.Completion.create(
|
||||
model=model_name,
|
||||
prompt=sample_chunks[0])
|
||||
|
||||
print(response)
|
||||
@@ -1,98 +0,0 @@
|
||||
import json
|
||||
import yt_dlp as youtube_dl
|
||||
from whisper_jax import FlaxWhisperPipline
|
||||
import jax.numpy as jnp
|
||||
|
||||
# Function to extract chapter information from a YouTube video URL
|
||||
def get_youtube_chapters(video_id):
|
||||
video_url = "https://www.youtube.com/watch?v=" + video_id
|
||||
ydl_opts = {
|
||||
'extract_flat': 'in_playlist',
|
||||
'skip_download': True,
|
||||
'quiet': True,
|
||||
}
|
||||
|
||||
with youtube_dl.YoutubeDL(ydl_opts) as ydl:
|
||||
video_info = ydl.extract_info(video_url, download=False)
|
||||
|
||||
chapters = []
|
||||
|
||||
if 'chapters' in video_info:
|
||||
for chapter in video_info['chapters']:
|
||||
start_time = chapter['start_time']
|
||||
end_time = chapter['end_time']
|
||||
title = chapter['title']
|
||||
|
||||
chapters.append({
|
||||
'start': start_time,
|
||||
'end': end_time,
|
||||
'title': title
|
||||
})
|
||||
|
||||
return chapters
|
||||
|
||||
|
||||
# Function to extract video transcription using yt_dlp
|
||||
def get_youtube_transcription(video_id):
|
||||
ydl_opts = {
|
||||
'format': 'bestaudio/best',
|
||||
'postprocessors': [{
|
||||
'key': 'FFmpegExtractAudio',
|
||||
'preferredcodec': 'mp3',
|
||||
'preferredquality': '192',
|
||||
}],
|
||||
'outtmpl': './artefacts/audio', # Specify output file path and name
|
||||
}
|
||||
|
||||
# Download the audio
|
||||
with youtube_dl.YoutubeDL(ydl_opts) as ydl:
|
||||
ydl.download(["https://www.youtube.com/watch?v=" + video_id])
|
||||
media_file = "./artefacts/audio.mp3"
|
||||
|
||||
pipeline = FlaxWhisperPipline("openai/whisper-" + "tiny",
|
||||
dtype=jnp.float16,
|
||||
batch_size=16)
|
||||
whisper_result = pipeline(media_file, return_timestamps=True)
|
||||
return whisper_result["chunks"]
|
||||
|
||||
|
||||
|
||||
# Function to scrape YouTube video transcripts and chapter information
|
||||
def scrape_youtube_data(video_id):
|
||||
transcript_text = get_youtube_transcription(video_id)
|
||||
chapters = get_youtube_chapters(video_id)
|
||||
print("transcript_text", transcript_text)
|
||||
print("chapters", chapters)
|
||||
return transcript_text, chapters
|
||||
|
||||
|
||||
# Function to generate fine-tuning dataset from YouTube data
|
||||
def generate_finetuning_dataset(video_ids):
|
||||
prompt_completion_pairs = []
|
||||
for video_id in video_ids:
|
||||
transcript_text, chapters = scrape_youtube_data(video_id)
|
||||
if transcript_text is not None and chapters is not None:
|
||||
for chapter in chapters:
|
||||
start_time = chapter["start"]
|
||||
end_time = chapter["end"]
|
||||
chapter_text = chapter["title"]
|
||||
|
||||
prompt = ""
|
||||
for transcript in transcript_text:
|
||||
if transcript["timestamp"][0] >= start_time and transcript["timestamp"][1] < end_time:
|
||||
prompt += transcript["text"]
|
||||
|
||||
if prompt is not None:
|
||||
completion = chapter_text
|
||||
prompt_completion_pairs.append({"prompt": prompt, "completion": completion})
|
||||
|
||||
return prompt_completion_pairs
|
||||
|
||||
|
||||
# Add all the video ids here, the videos must have captions [chapters]
|
||||
video_ids = ["yTnSEZIwnkU"]
|
||||
dataset = generate_finetuning_dataset(video_ids)
|
||||
|
||||
with open("finetuning_dataset.jsonl", "w", encoding="utf-8") as file:
|
||||
for example in dataset:
|
||||
file.write(json.dumps(example) + "\n")
|
||||
@@ -1,188 +0,0 @@
|
||||
import asyncio
|
||||
import datetime
|
||||
import io
|
||||
import json
|
||||
import threading
|
||||
import uuid
|
||||
import wave
|
||||
from concurrent.futures import ThreadPoolExecutor
|
||||
|
||||
import jax.numpy as jnp
|
||||
import requests
|
||||
from aiohttp import web
|
||||
from aiortc import MediaStreamTrack, RTCPeerConnection, RTCSessionDescription
|
||||
from aiortc.contrib.media import MediaRelay
|
||||
from av import AudioFifo
|
||||
from sortedcontainers import SortedDict
|
||||
from whisper_jax import FlaxWhisperPipline
|
||||
|
||||
from reflector.utils.log_utils import LOGGER
|
||||
from reflector.utils.run_utils import CONFIG, Mutex
|
||||
|
||||
WHISPER_MODEL_SIZE = CONFIG['WHISPER']["WHISPER_REAL_TIME_MODEL_SIZE"]
|
||||
pcs = set()
|
||||
relay = MediaRelay()
|
||||
data_channel = None
|
||||
sorted_message_queue = SortedDict()
|
||||
CHANNELS = 2
|
||||
RATE = 44100
|
||||
CHUNK_SIZE = 256
|
||||
pipeline = FlaxWhisperPipline("openai/whisper-" + WHISPER_MODEL_SIZE,
|
||||
dtype=jnp.float16,
|
||||
batch_size=16)
|
||||
start_time = datetime.datetime.now()
|
||||
executor = ThreadPoolExecutor()
|
||||
audio_buffer = AudioFifo()
|
||||
frame_lock = Mutex(audio_buffer)
|
||||
|
||||
|
||||
def channel_log(channel, t, message):
|
||||
print("channel(%s) %s %s" % (channel.label, t, message))
|
||||
|
||||
|
||||
def thread_queue_channel_send():
|
||||
loop = asyncio.new_event_loop()
|
||||
asyncio.set_event_loop(loop)
|
||||
try:
|
||||
least_time = sorted_message_queue.keys()[0]
|
||||
message = sorted_message_queue[least_time]
|
||||
if message:
|
||||
del sorted_message_queue[least_time]
|
||||
data_channel.send(message)
|
||||
except Exception as e:
|
||||
print("Exception", str(e))
|
||||
pass
|
||||
loop.run_forever()
|
||||
|
||||
|
||||
def get_transcription():
|
||||
while True:
|
||||
with frame_lock.lock() as audio_buffer:
|
||||
frames = audio_buffer.read_many(CHUNK_SIZE * 960, partial=False)
|
||||
if not frames:
|
||||
transcribe = False
|
||||
else:
|
||||
transcribe = True
|
||||
|
||||
if transcribe:
|
||||
print("Transcribing..")
|
||||
try:
|
||||
sorted_message_queue[frames[0].time] = None
|
||||
out_file = io.BytesIO()
|
||||
wf = wave.open(out_file, "wb")
|
||||
wf.setnchannels(CHANNELS)
|
||||
wf.setframerate(RATE)
|
||||
wf.setsampwidth(2)
|
||||
|
||||
for frame in frames:
|
||||
wf.writeframes(b''.join(frame.to_ndarray()))
|
||||
wf.close()
|
||||
|
||||
whisper_result = pipeline(out_file.getvalue())
|
||||
item = {
|
||||
'text': whisper_result["text"],
|
||||
'start_time': str(frames[0].time),
|
||||
'time': str(datetime.datetime.now())
|
||||
}
|
||||
sorted_message_queue[frames[0].time] = str(item)
|
||||
start_messaging_thread()
|
||||
except Exception as e:
|
||||
print("Exception -> ", str(e))
|
||||
|
||||
|
||||
class AudioStreamTrack(MediaStreamTrack):
|
||||
"""
|
||||
An audio stream track to send audio frames.
|
||||
"""
|
||||
|
||||
kind = "audio"
|
||||
|
||||
def __init__(self, track):
|
||||
super().__init__() # don't forget this!
|
||||
self.track = track
|
||||
|
||||
async def recv(self):
|
||||
frame = await self.track.recv()
|
||||
audio_buffer.write(frame)
|
||||
return frame
|
||||
|
||||
|
||||
def start_messaging_thread():
|
||||
message_thread = threading.Thread(target=thread_queue_channel_send)
|
||||
message_thread.start()
|
||||
|
||||
|
||||
def start_transcription_thread(max_threads: int):
|
||||
for i in range(max_threads):
|
||||
t_thread = threading.Thread(target=get_transcription)
|
||||
t_thread.start()
|
||||
|
||||
|
||||
async def offer(request: requests.Request):
|
||||
params = await request.json()
|
||||
offer = RTCSessionDescription(sdp=params["sdp"], type=params["type"])
|
||||
|
||||
pc = RTCPeerConnection()
|
||||
pc_id = "PeerConnection(%s)" % uuid.uuid4()
|
||||
pcs.add(pc)
|
||||
|
||||
def log_info(msg: str, *args):
|
||||
LOGGER.info(pc_id + " " + msg, *args)
|
||||
|
||||
log_info("Created for " + request.remote)
|
||||
|
||||
@pc.on("datachannel")
|
||||
def on_datachannel(channel):
|
||||
global data_channel, start_time
|
||||
data_channel = channel
|
||||
channel_log(channel, "-", "created by remote party")
|
||||
start_time = datetime.datetime.now()
|
||||
|
||||
@channel.on("message")
|
||||
def on_message(message: str):
|
||||
channel_log(channel, "<", message)
|
||||
if isinstance(message, str) and message.startswith("ping"):
|
||||
# reply
|
||||
channel.send("pong" + message[4:])
|
||||
|
||||
@pc.on("connectionstatechange")
|
||||
async def on_connectionstatechange():
|
||||
log_info("Connection state is " + pc.connectionState)
|
||||
if pc.connectionState == "failed":
|
||||
await pc.close()
|
||||
pcs.discard(pc)
|
||||
|
||||
@pc.on("track")
|
||||
def on_track(track):
|
||||
log_info("Track " + track.kind + " received")
|
||||
pc.addTrack(AudioStreamTrack(relay.subscribe(track)))
|
||||
|
||||
# handle offer
|
||||
await pc.setRemoteDescription(offer)
|
||||
|
||||
# send answer
|
||||
answer = await pc.createAnswer()
|
||||
await pc.setLocalDescription(answer)
|
||||
return web.Response(
|
||||
content_type="application/json",
|
||||
text=json.dumps({
|
||||
"sdp": pc.localDescription.sdp,
|
||||
"type": pc.localDescription.type
|
||||
}),
|
||||
)
|
||||
|
||||
|
||||
async def on_shutdown(app: web.Application):
|
||||
coros = [pc.close() for pc in pcs]
|
||||
await asyncio.gather(*coros)
|
||||
pcs.clear()
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
app = web.Application()
|
||||
app.on_shutdown.append(on_shutdown)
|
||||
start_transcription_thread(6)
|
||||
app.router.add_post("/offer", offer)
|
||||
web.run_app(
|
||||
app, access_log=None, host="127.0.0.1", port=1250
|
||||
)
|
||||
@@ -1,57 +0,0 @@
|
||||
import requests
|
||||
import spacy
|
||||
|
||||
# Enter the Machine where the LLM is hosted
|
||||
LLM_MACHINE_IP = ""
|
||||
# This is the URL of text-generation-webui
|
||||
URL = f"http://{LLM_MACHINE_IP}:5000/api/v1/generate"
|
||||
|
||||
headers = {
|
||||
"Content-Type": "application/json"
|
||||
}
|
||||
|
||||
|
||||
def split_text_file(filename, token_count):
|
||||
nlp = spacy.load('en_core_web_md')
|
||||
|
||||
with open(filename, 'r') as file:
|
||||
text = file.read()
|
||||
|
||||
doc = nlp(text)
|
||||
total_tokens = len(doc)
|
||||
|
||||
parts = []
|
||||
start_index = 0
|
||||
|
||||
while start_index < total_tokens:
|
||||
end_index = start_index + token_count
|
||||
part_tokens = doc[start_index:end_index - 5]
|
||||
part = ' '.join(token.text for token in part_tokens)
|
||||
parts.append(part)
|
||||
start_index = end_index
|
||||
|
||||
return parts
|
||||
|
||||
|
||||
final_summary = ""
|
||||
parts = split_text_file("transcript.txt", 1600)
|
||||
|
||||
for part in parts:
|
||||
prompt = f"""
|
||||
### Human:
|
||||
Given the following text, distill the most important information
|
||||
into a short summary: {part}
|
||||
|
||||
### Assistant:
|
||||
"""
|
||||
data = {
|
||||
"prompt": prompt
|
||||
}
|
||||
try:
|
||||
response = requests.post(URL, headers=headers, json=data)
|
||||
print(response.json())
|
||||
except Exception as e:
|
||||
print(str(e))
|
||||
|
||||
with open("summary.txt", "w") as sum:
|
||||
sum.write(" ".join(final_summary))
|
||||
@@ -1,43 +0,0 @@
|
||||
import torch
|
||||
from transformers import BertTokenizer, BertModel
|
||||
from sentence_transformers import SentenceTransformer
|
||||
from sklearn.metrics.pairwise import cosine_similarity
|
||||
|
||||
# Load the pre-trained BERT model and tokenizer
|
||||
model_name = "bert-base-uncased"
|
||||
model = BertModel.from_pretrained(model_name)
|
||||
tokenizer = BertTokenizer.from_pretrained(model_name)
|
||||
|
||||
# Set the device to use
|
||||
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
||||
model.to(device)
|
||||
|
||||
# Load the SentenceTransformer model
|
||||
sentence_transformer_model = SentenceTransformer('average_word_embeddings_glove.6B.300d')
|
||||
|
||||
# Define the input text
|
||||
text = "Your input text to be summarized goes here."
|
||||
|
||||
# Tokenize the text
|
||||
tokens = tokenizer.tokenize(text)
|
||||
input_ids = tokenizer.convert_tokens_to_ids(tokens)
|
||||
input_ids = torch.tensor([input_ids]).to(device)
|
||||
|
||||
# Get the BERT model output
|
||||
with torch.no_grad():
|
||||
outputs = model(input_ids)[0] # Extract the last hidden states
|
||||
|
||||
# Calculate sentence embeddings
|
||||
sentence_embeddings = outputs.mean(dim=1).squeeze().cpu().numpy()
|
||||
input_text_embedding = sentence_transformer_model.encode([text])[0]
|
||||
|
||||
# Calculate cosine similarity between sentences and input text
|
||||
similarity_scores = cosine_similarity([input_text_embedding], sentence_embeddings)
|
||||
|
||||
# Sort the sentences by similarity scores in descending order
|
||||
sorted_sentences = [sent for _, sent in sorted(zip(similarity_scores[0], sentences), reverse=True)]
|
||||
|
||||
# Choose the top sentences as the summary
|
||||
num_summary_sentences = 2 # Adjust as needed
|
||||
summary = ". ".join(sorted_sentences[:num_summary_sentences])
|
||||
print("Summary:", summary)
|
||||
@@ -1,79 +0,0 @@
|
||||
"""
|
||||
This is an example code containing the bare essentials to load a chat
|
||||
LLM and infer from it using a predefined prompt. The purpose of this file
|
||||
is to show an example of inferring from a chat LLM which is required for
|
||||
banana.dev due to its design and platform limitations
|
||||
"""
|
||||
|
||||
# The following logic was tested on the monadical-ml machine
|
||||
|
||||
import json
|
||||
|
||||
import torch
|
||||
from transformers import (
|
||||
AutoModelForCausalLM,
|
||||
AutoTokenizer
|
||||
)
|
||||
from transformers.generation import GenerationConfig
|
||||
|
||||
# This can be passed via the environment variable or the params supplied
|
||||
# when starting the program via banana.dev platform
|
||||
MODEL_NAME = "lmsys/vicuna-13b-v1.5"
|
||||
|
||||
# Load the model in half precision, and less memory usage
|
||||
model = AutoModelForCausalLM.from_pretrained(MODEL_NAME,
|
||||
low_cpu_mem_usage=True,
|
||||
torch_dtype=torch.bfloat16
|
||||
)
|
||||
|
||||
# Generation config
|
||||
model.config.max_new_tokens = 300
|
||||
gen_cfg = GenerationConfig.from_model_config(model.config)
|
||||
gen_cfg.max_new_tokens = 300
|
||||
|
||||
# Load the tokenizer
|
||||
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
|
||||
|
||||
# Move model to GPU
|
||||
model = model.cuda()
|
||||
print(f"Loading {MODEL_NAME} successful")
|
||||
|
||||
# Inputs
|
||||
sample_chunks = [
|
||||
"You all just came off of your incredible Google Cloud next conference where you released a wide variety of functionality and features and new products across artisan television and also across the entire sort of cloud ecosystem . You want to just first by walking through , first start by walking through all the innovations that you sort of released and what you 're excited about when you come to Google Cloud ? Now our vision is super simple . If you look at what smartphones did for a consumer , you know they took a computer and internet browser , a communication device , and a camera , and made it so that it 's in everybody 's pocket , so it really brought computation to every person . We feel that , you know , our , what we 're trying to do is take all the technological innovation that Google 's doing , but make it super simple so that everyone can consume it . And so that includes our global data center footprint , all the new types of hardware and large-scale systems we work on , the software that we 're making available for people to do high-scale computation , tools for data processing , tools for cybersecurity , processing , tools for cyber security , tools for machine learning , but make it so simple that everyone can use it . And every step that we do to simplify things for people , we think adoption can grow . And so that 's a lot of what we 've done these last three , four years , and we made a number of announcements that next in machine learning and AI in particular , you know , we look at our work as four elements , how we take our large-scale compute systems that were building for AI and how we make that available to everybody . Second , what we 're doing with the software stacks and top of it , things like jacks and other things and how we 're making those available to everybody . Third is advances because different people have different levels of expertise . Some people say I need the hardware to build my own large language model or algorithm . Other people say , look , I really need to use a building block . You guys give me . So , 30s we 've done a lot with AutoML and we announce new capability for image , video , and translation to make it available to everybody . And then lastly , we 're also building completely packaged solutions for some areas and we announce some new stuff . ",
|
||||
" We 're joined next by Thomas Curian , CEO of Google Cloud , and Alexander Wang , CEO and founder of Scale AI . Thomas joined Google in November 2018 as the CEO of Google Cloud . Prior to Google , Thomas spent 22 years at Oracle , where most recently he was president of product development . Before that , Thomas worked at McKinsey as a business analyst and engagement manager . His nearly 30 years of experience have given him a deep knowledge of engineering enterprise relationships and leadership of large organizations . Thomas 's degrees include an MBA in administration and management from Stanford University , as an RJ Miller scholar and a BSEE in electrical engineering and computer science from Princeton University , where he graduated suma cum laude . Thomas serves as a member of the Stanford graduate School of Business Advisory Council and Princeton University School of Engineering Advisory Council . Please welcome to the stage , Thomas Curian and Alexander Wang . This is a super exciting conversation . Thanks for being here , Thomas ."]
|
||||
|
||||
# Model Prompt template for current model
|
||||
prompt = f"""
|
||||
### Human:
|
||||
Create a JSON object as response.The JSON object must have 2 fields:
|
||||
i) title and ii) summary.For the title field,generate a short title
|
||||
for the given text. For the summary field, summarize the given text
|
||||
in three sentences.
|
||||
|
||||
{sample_chunks[0]}
|
||||
|
||||
### Assistant:
|
||||
"""
|
||||
|
||||
# Inference : Chat generation
|
||||
input_ids = tokenizer.encode(prompt, return_tensors='pt').to(model.device)
|
||||
output = model.generate(input_ids, generation_config=gen_cfg)
|
||||
|
||||
# Process output
|
||||
response = tokenizer.decode(output[0].cpu(), skip_special_tokens=True)
|
||||
response = response.split("### Assistant:\n")
|
||||
print("TitleSummaryJsonResponse :", json.loads(response[1]))
|
||||
print("Inference successful")
|
||||
|
||||
# Sample response for sample_chunks[0]
|
||||
|
||||
# TitleSummaryJsonResponse :
|
||||
# {
|
||||
# 'title': 'Google Cloud Next Conference: Simplifying AI and Machine Learning for Everyone',
|
||||
# 'summary': 'Google Cloud announced a wide range of innovations and new products in the AI
|
||||
# and machine learning space at the recent Google Cloud Next conference. The goal
|
||||
# is to make these technologies accessible to everyone by simplifying the process
|
||||
# and providing tools for data processing, cybersecurity, and machine learning.
|
||||
# Google is also working on advances in AutoML and packaged solutions for certain areas.'
|
||||
# }
|
||||
@@ -1,101 +0,0 @@
|
||||
# Approach 1
|
||||
from transformers import GPTNeoForCausalLM, GPT2Tokenizer
|
||||
|
||||
model_name = 'EleutherAI/gpt-neo-1.3B'
|
||||
tokenizer = GPT2Tokenizer.from_pretrained(model_name)
|
||||
model = GPTNeoForCausalLM.from_pretrained(model_name)
|
||||
|
||||
conversation = """
|
||||
Summarize the following conversation in 3 key sentences:
|
||||
|
||||
We 're joined next by Thomas Curian , CEO of Google Cloud , and Alexander Wang , CEO and founder of Scale AI .
|
||||
Thomas joined Google in November 2018 as the CEO of Google Cloud . Prior to Google , Thomas spent 22 years at Oracle , where most recently he was president of product development .
|
||||
Before that , Thomas worked at McKinsey as a business analyst and engagement manager . His nearly 30 years of experience have given him a deep knowledge of engineering enterprise relationships and leadership of large organizations .
|
||||
Thomas 's degrees include an MBA in administration and management from Stanford University , as an RJ Miller scholar and a BSEE in electrical engineering and computer science from Princeton University , where he graduated suma cum laude .
|
||||
Thomas serves as a member of the Stanford graduate School of Business Advisory Council and Princeton University School of Engineering Advisory Council .
|
||||
Please welcome to the stage , Thomas Curian and Alexander Wang . This is a super exciting conversation . Thanks for being here , Thomas .
|
||||
"""
|
||||
|
||||
input_ids = tokenizer.encode(conversation, return_tensors='pt')
|
||||
|
||||
output = model.generate(input_ids,
|
||||
max_length=30,
|
||||
num_return_sequences=1)
|
||||
|
||||
caption = tokenizer.decode(output[0], skip_special_tokens=True)
|
||||
print("Caption:", caption[len(input_ids):])
|
||||
|
||||
|
||||
# Approach 2
|
||||
import torch
|
||||
from transformers import GPT2LMHeadModel, GPT2Tokenizer
|
||||
|
||||
model_name = "gpt2"
|
||||
tokenizer = GPT2Tokenizer.from_pretrained(model_name)
|
||||
model = GPT2LMHeadModel.from_pretrained(model_name)
|
||||
|
||||
model.eval()
|
||||
|
||||
text = """
|
||||
You all just came off of your incredible Google Cloud next conference where you released a wide variety of functionality and features and new products across artisan television and also across the entire sort of cloud ecosystem . You want to just first by walking through , first start by walking through all the innovations that you sort of released and what you 're excited about when you come to Google Cloud ? Now our vision is super simple . If you look at what smartphones did for a consumer , you know they took a computer and internet browser , a communication device , and a camera , and made it so that it 's in everybody 's pocket , so it really brought computation to every person . We feel that , you know , our , what we 're trying to do is take all the technological innovation that Google 's doing , but make it super simple so that everyone can consume it . And so that includes our global data center footprint , all the new types of hardware and large-scale systems we work on , the software that we 're making available for people to do high-scale computation , tools for data processing , tools for cybersecurity , processing , tools for cyber security , tools for machine learning , but make it so simple that everyone can use it . And every step that we do to simplify things for people , we think adoption can grow . And so that 's a lot of what we 've done these last three , four years , and we made a number of announcements that next in machine learning and AI in particular , you know , we look at our work as four elements , how we take our large-scale compute systems that were building for AI and how we make that available to everybody . Second , what we 're doing with the software stacks and top of it , things like jacks and other things and how we 're making those available to everybody . Third is advances because different people have different levels of expertise . Some people say I need the hardware to build my own large language model or algorithm . Other people say , look , I really need to use a building block . You guys give me . So , 30s we 've done a lot with AutoML and we announce new capability for image , video , and translation to make it available to everybody . And then lastly , we 're also building completely packaged solutions for some areas and we announce some new stuff . "
|
||||
"""
|
||||
|
||||
tokenizer.pad_token = tokenizer.eos_token
|
||||
input_ids = tokenizer.encode(text,
|
||||
max_length=100,
|
||||
truncation=True,
|
||||
return_tensors="pt")
|
||||
attention_mask = torch.ones(input_ids.shape, dtype=torch.long)
|
||||
output = model.generate(input_ids,
|
||||
max_new_tokens=20,
|
||||
num_return_sequences=1,
|
||||
num_beams=2,
|
||||
attention_mask=attention_mask)
|
||||
|
||||
chapter_titles = [tokenizer.decode(output[i], skip_special_tokens=True) for i in range(output.shape[0])]
|
||||
for i, title in enumerate(chapter_titles):
|
||||
print("Caption: ", title)
|
||||
|
||||
# Approach 3
|
||||
|
||||
import torch
|
||||
from transformers import GPT2LMHeadModel, GPT2Tokenizer
|
||||
|
||||
|
||||
def generate_response(conversation, max_length=100):
|
||||
input_text = ""
|
||||
for entry in conversation:
|
||||
role = entry["role"]
|
||||
content = entry["content"]
|
||||
input_text += f"{role}: {content}\n"
|
||||
|
||||
# Tokenize the entire conversation
|
||||
input_ids = tokenizer.encode(input_text, return_tensors="pt")
|
||||
|
||||
# Generate text based on the entire conversation
|
||||
with torch.no_grad():
|
||||
output = model.generate(input_ids, pad_token_id=tokenizer.eos_token_id)
|
||||
|
||||
# Decode the generated text and return it
|
||||
response = tokenizer.decode(output[0], skip_special_tokens=True)
|
||||
return response
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
|
||||
# Call appropriate approach from the main while experimenting
|
||||
model_name = "gpt2"
|
||||
model = GPT2LMHeadModel.from_pretrained(model_name)
|
||||
tokenizer = GPT2Tokenizer.from_pretrained(model_name)
|
||||
|
||||
sample_chunks = [
|
||||
"You all just came off of your incredible Google Cloud next conference where you released a wide variety of functionality and features and new products across artisan television and also across the entire sort of cloud ecosystem . You want to just first by walking through , first start by walking through all the innovations that you sort of released and what you 're excited about when you come to Google Cloud ? Now our vision is super simple . If you look at what smartphones did for a consumer , you know they took a computer and internet browser , a communication device , and a camera , and made it so that it 's in everybody 's pocket , so it really brought computation to every person . We feel that , you know , our , what we 're trying to do is take all the technological innovation that Google 's doing , but make it super simple so that everyone can consume it . And so that includes our global data center footprint , all the new types of hardware and large-scale systems we work on , the software that we 're making available for people to do high-scale computation , tools for data processing , tools for cybersecurity , processing , tools for cyber security , tools for machine learning , but make it so simple that everyone can use it . And every step that we do to simplify things for people , we think adoption can grow . And so that 's a lot of what we 've done these last three , four years , and we made a number of announcements that next in machine learning and AI in particular , you know , we look at our work as four elements , how we take our large-scale compute systems that were building for AI and how we make that available to everybody . Second , what we 're doing with the software stacks and top of it , things like jacks and other things and how we 're making those available to everybody . Third is advances because different people have different levels of expertise . Some people say I need the hardware to build my own large language model or algorithm . Other people say , look , I really need to use a building block . You guys give me . So , 30s we 've done a lot with AutoML and we announce new capability for image , video , and translation to make it available to everybody . And then lastly , we 're also building completely packaged solutions for some areas and we announce some new stuff . "
|
||||
]
|
||||
|
||||
conversation = [
|
||||
{"role": "system", "content": "Summarize this text"},
|
||||
{"role": "user", "content": " text : " + sample_chunks[0]},
|
||||
]
|
||||
|
||||
response = generate_response(conversation)
|
||||
print("Response:", response)
|
||||
@@ -1,157 +0,0 @@
|
||||
import spacy
|
||||
import sys
|
||||
|
||||
|
||||
# Observe the incremental summaries by performing summaries in chunks
|
||||
with open("transcript.txt", "r", encoding="utf-8") as file:
|
||||
transcription = file.read()
|
||||
|
||||
|
||||
def split_text_file(filename, token_count):
|
||||
nlp = spacy.load('en_core_web_md')
|
||||
|
||||
with open(filename, 'r', encoding="utf-8") as file:
|
||||
text = file.read()
|
||||
|
||||
doc = nlp(text)
|
||||
total_tokens = len(doc)
|
||||
|
||||
parts = []
|
||||
start_index = 0
|
||||
|
||||
while start_index < total_tokens:
|
||||
end_index = start_index + token_count
|
||||
part_tokens = doc[start_index:end_index]
|
||||
part = ' '.join(token.text for token in part_tokens)
|
||||
parts.append(part)
|
||||
start_index = end_index
|
||||
|
||||
return parts
|
||||
|
||||
|
||||
# Set the chunk length here to split the transcript and test
|
||||
MAX_CHUNK_LENGTH = 1000
|
||||
|
||||
chunks = split_text_file("transcript.txt", MAX_CHUNK_LENGTH)
|
||||
print("Number of chunks", len(chunks))
|
||||
|
||||
# Write chunks to file to refer to input vs output, separated by blank lines
|
||||
with open("chunks" + str(MAX_CHUNK_LENGTH) + ".txt", "a", encoding="utf-8") as file:
|
||||
for c in chunks:
|
||||
file.write(c + "\n\n")
|
||||
|
||||
# If we want to run only a certain model, type the option while running
|
||||
# ex. python incsum.py 1 => will run approach 1
|
||||
# If no input, will run all approaches
|
||||
|
||||
try:
|
||||
index = sys.argv[1]
|
||||
except:
|
||||
index = None
|
||||
|
||||
# Approach 1 : facebook/bart-large-cnn
|
||||
if index == "1" or index is None:
|
||||
SUMMARY_MODEL = "facebook/bart-large-cnn"
|
||||
MIN_LENGTH = 5
|
||||
MAX_LENGTH = 10
|
||||
BEAM_SIZE = 2
|
||||
|
||||
print("Performing chunk summary : " + SUMMARY_MODEL)
|
||||
|
||||
from transformers import BartTokenizer, BartForConditionalGeneration
|
||||
|
||||
tokenizer = BartTokenizer.from_pretrained(SUMMARY_MODEL)
|
||||
model = BartForConditionalGeneration.from_pretrained(SUMMARY_MODEL)
|
||||
summaries = []
|
||||
for c in chunks:
|
||||
input_ids = tokenizer.encode(c,
|
||||
truncation=True,
|
||||
max_length=MAX_CHUNK_LENGTH,
|
||||
padding="max_length",
|
||||
return_tensors='pt')
|
||||
summary_ids = model.generate(
|
||||
input_ids,
|
||||
num_beams=BEAM_SIZE,
|
||||
max_length=56,
|
||||
early_stopping=True,
|
||||
length_penalty=1.0)
|
||||
summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)
|
||||
summaries.append(summary)
|
||||
|
||||
with open("bart-summaries.txt", "a", encoding="utf-8") as file:
|
||||
for summary in summaries:
|
||||
file.write(summary + "\n\n")
|
||||
|
||||
# Approach 2
|
||||
if index == "2" or index is None:
|
||||
print("Performing chunk summary : " + "gpt-neo-1.3B")
|
||||
|
||||
import torch
|
||||
from transformers import GPTNeoForCausalLM, GPT2Tokenizer
|
||||
|
||||
model = GPTNeoForCausalLM.from_pretrained("EleutherAI/gpt-neo-1.3B")
|
||||
tokenizer = GPT2Tokenizer.from_pretrained("EleutherAI/gpt-neo-1.3B")
|
||||
tokenizer.add_special_tokens({'pad_token': '[PAD]'})
|
||||
summaries = []
|
||||
|
||||
for c in chunks:
|
||||
input_ids = tokenizer.encode(c,
|
||||
truncation=True,
|
||||
return_tensors='pt')
|
||||
input_length = input_ids.shape[1]
|
||||
attention_mask = torch.ones(input_ids.shape, dtype=torch.long)
|
||||
|
||||
max_summary_length = 100
|
||||
max_length = input_length + max_summary_length
|
||||
|
||||
output = model.generate(input_ids,
|
||||
max_length=max_length,
|
||||
attention_mask=attention_mask,
|
||||
pad_token_id=model.config.eos_token_id,
|
||||
num_beams=4,
|
||||
length_penalty=2.0,
|
||||
early_stopping=True)
|
||||
summary_ids = output[0, input_length:]
|
||||
summary = tokenizer.decode(summary_ids, skip_special_tokens=True)
|
||||
summaries.append(summary)
|
||||
with open("gptneo1.3B-summaries.txt", "a", encoding="utf-8") as file:
|
||||
file.write(summary + "\n\n")
|
||||
|
||||
# Approach 3
|
||||
if index == "3" or index is None:
|
||||
print("Performing chunk summary : " + "mpt-7B")
|
||||
|
||||
import torch
|
||||
import transformers
|
||||
from transformers import AutoTokenizer
|
||||
|
||||
config = transformers.AutoConfig.from_pretrained('mosaicml/mpt-7b',
|
||||
trust_remote_code=True)
|
||||
config.attn_config['attn_impl'] = 'triton'
|
||||
config.max_seq_len = 1024
|
||||
config.init_device = "meta"
|
||||
|
||||
model = transformers.AutoModelForCausalLM.from_pretrained(
|
||||
'mosaicml/mpt-7b',
|
||||
trust_remote_code=True,
|
||||
torch_dtype=torch.bfloat16
|
||||
)
|
||||
|
||||
tokenizer = AutoTokenizer.from_pretrained('EleutherAI/gpt-neox-20b')
|
||||
|
||||
summaries = []
|
||||
for c in chunks:
|
||||
input_ids = tokenizer.encode(c, return_tensors="pt")
|
||||
attention_mask = torch.ones(input_ids.shape, dtype=torch.long)
|
||||
output = model.generate(input_ids,
|
||||
max_new_tokens=25,
|
||||
attention_mask=attention_mask,
|
||||
pad_token_id=model.config.eos_token_id,
|
||||
num_return_sequences=1)
|
||||
summary = tokenizer.decode(output[0],
|
||||
skip_special_tokens=True)
|
||||
summaries.append(summary)
|
||||
|
||||
with open("mpt-7b-summaries.txt", "a", encoding="utf-8") as file:
|
||||
for summary in summaries:
|
||||
file.write(summary + "\n\n")
|
||||
@@ -1,37 +0,0 @@
|
||||
# Use OpenAI API endpoint to send data to OpenAI
|
||||
# along with prompts to caption/summarize the conversation
|
||||
|
||||
import openai
|
||||
|
||||
openai.api_key = ""
|
||||
|
||||
# to caption, user prompt used : "caption this conversation"
|
||||
# max_tokens=20
|
||||
|
||||
# to incremental summarize, user prompt used : "summarize this conversation in a few sentences by taking key points"
|
||||
# max_tokens=300
|
||||
|
||||
sample_chunks = [
|
||||
"You all just came off of your incredible Google Cloud next conference where you released a wide variety of functionality and features and new products across artisan television and also across the entire sort of cloud ecosystem . You want to just first by walking through , first start by walking through all the innovations that you sort of released and what you 're excited about when you come to Google Cloud ? Now our vision is super simple . If you look at what smartphones did for a consumer , you know they took a computer and internet browser , a communication device , and a camera , and made it so that it 's in everybody 's pocket , so it really brought computation to every person . We feel that , you know , our , what we 're trying to do is take all the technological innovation that Google 's doing , but make it super simple so that everyone can consume it . And so that includes our global data center footprint , all the new types of hardware and large-scale systems we work on , the software that we 're making available for people to do high-scale computation , tools for data processing , tools for cybersecurity , processing , tools for cyber security , tools for machine learning , but make it so simple that everyone can use it . And every step that we do to simplify things for people , we think adoption can grow . And so that 's a lot of what we 've done these last three , four years , and we made a number of announcements that next in machine learning and AI in particular , you know , we look at our work as four elements , how we take our large-scale compute systems that were building for AI and how we make that available to everybody . Second , what we 're doing with the software stacks and top of it , things like jacks and other things and how we 're making those available to everybody . Third is advances because different people have different levels of expertise . Some people say I need the hardware to build my own large language model or algorithm . Other people say , look , I really need to use a building block . You guys give me . So , 30s we 've done a lot with AutoML and we announce new capability for image , video , and translation to make it available to everybody . And then lastly , we 're also building completely packaged solutions for some areas and we announce some new stuff . ",
|
||||
" We 're joined next by Thomas Curian , CEO of Google Cloud , and Alexander Wang , CEO and founder of Scale AI . Thomas joined Google in November 2018 as the CEO of Google Cloud . Prior to Google , Thomas spent 22 years at Oracle , where most recently he was president of product development . Before that , Thomas worked at McKinsey as a business analyst and engagement manager . His nearly 30 years of experience have given him a deep knowledge of engineering enterprise relationships and leadership of large organizations . Thomas 's degrees include an MBA in administration and management from Stanford University , as an RJ Miller scholar and a BSEE in electrical engineering and computer science from Princeton University , where he graduated suma cum laude . Thomas serves as a member of the Stanford graduate School of Business Advisory Council and Princeton University School of Engineering Advisory Council . Please welcome to the stage , Thomas Curian and Alexander Wang . This is a super exciting conversation . Thanks for being here , Thomas ."]
|
||||
|
||||
conversation = [
|
||||
{"role": "system",
|
||||
"content": sample_chunks[1]},
|
||||
{"role": "user",
|
||||
"content": "summarize this conversation in a few sentences by taking key points"}
|
||||
]
|
||||
|
||||
model = "gpt-3.5-turbo"
|
||||
response = openai.ChatCompletion.create(model=model,
|
||||
messages=conversation,
|
||||
n=1,
|
||||
max_tokens=300)
|
||||
|
||||
# Try fine tuned model
|
||||
# model = "davinci:ft-personal-2023-07-14-10-43-51"
|
||||
# response = openai.Completion.create(model=model,
|
||||
# prompt=sample_chunks[0] + " -> ")
|
||||
|
||||
caption = response.choices[0]
|
||||
print(caption)
|
||||
@@ -1,33 +0,0 @@
|
||||
from transformers import PegasusForConditionalGeneration, PegasusTokenizer
|
||||
import torch
|
||||
# Load the Pegasus model and tokenizer
|
||||
model_name = "google/pegasus-large"
|
||||
model = PegasusForConditionalGeneration.from_pretrained(model_name)
|
||||
tokenizer = PegasusTokenizer.from_pretrained(model_name)
|
||||
|
||||
# Set the device to use
|
||||
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
||||
model.to(device)
|
||||
|
||||
sample_chunks = ["You all just came off of your incredible Google Cloud next conference where you released a wide variety of functionality and features and new products across artisan television and also across the entire sort of cloud ecosystem . You want to just first by walking through , first start by walking through all the innovations that you sort of released and what you 're excited about when you come to Google Cloud ? Now our vision is super simple . If you look at what smartphones did for a consumer , you know they took a computer and internet browser , a communication device , and a camera , and made it so that it 's in everybody 's pocket , so it really brought computation to every person . We feel that , you know , our , what we 're trying to do is take all the technological innovation that Google 's doing , but make it super simple so that everyone can consume it . And so that includes our global data center footprint , all the new types of hardware and large-scale systems we work on , the software that we 're making available for people to do high-scale computation , tools for data processing , tools for cybersecurity , processing , tools for cyber security , tools for machine learning , but make it so simple that everyone can use it . And every step that we do to simplify things for people , we think adoption can grow . And so that 's a lot of what we 've done these last three , four years , and we made a number of announcements that next in machine learning and AI in particular , you know , we look at our work as four elements , how we take our large-scale compute systems that were building for AI and how we make that available to everybody . Second , what we 're doing with the software stacks and top of it , things like jacks and other things and how we 're making those available to everybody . Third is advances because different people have different levels of expertise . Some people say I need the hardware to build my own large language model or algorithm . Other people say , look , I really need to use a building block . You guys give me . So , 30s we 've done a lot with AutoML and we announce new capability for image , video , and translation to make it available to everybody . And then lastly , we 're also building completely packaged solutions for some areas and we announce some new stuff . ",
|
||||
" We 're joined next by Thomas Curian , CEO of Google Cloud , and Alexander Wang , CEO and founder of Scale AI . Thomas joined Google in November 2018 as the CEO of Google Cloud . Prior to Google , Thomas spent 22 years at Oracle , where most recently he was president of product development . Before that , Thomas worked at McKinsey as a business analyst and engagement manager . His nearly 30 years of experience have given him a deep knowledge of engineering enterprise relationships and leadership of large organizations . Thomas 's degrees include an MBA in administration and management from Stanford University , as an RJ Miller scholar and a BSEE in electrical engineering and computer science from Princeton University , where he graduated suma cum laude . Thomas serves as a member of the Stanford graduate School of Business Advisory Council and Princeton University School of Engineering Advisory Council . Please welcome to the stage , Thomas Curian and Alexander Wang . This is a super exciting conversation . Thanks for being here , Thomas ."]
|
||||
|
||||
|
||||
# Define the input text for summarization
|
||||
text = sample_chunks[1]
|
||||
|
||||
inputs = tokenizer(text, truncation=True, padding="longest", return_tensors="pt").to(device)
|
||||
|
||||
# Generate the summary
|
||||
summary_ids = model.generate(
|
||||
inputs["input_ids"],
|
||||
attention_mask=inputs["attention_mask"],
|
||||
max_length=200,
|
||||
num_beams=4,
|
||||
length_penalty=2.0,
|
||||
early_stopping=True,
|
||||
)
|
||||
|
||||
# Decode and print the summary
|
||||
summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)
|
||||
print("Summary:", summary)
|
||||
@@ -1,27 +0,0 @@
|
||||
from transformers import T5ForConditionalGeneration, T5Tokenizer
|
||||
import torch
|
||||
# Load the T5 model and tokenizer
|
||||
model_name = "t5-base"
|
||||
model = T5ForConditionalGeneration.from_pretrained(model_name)
|
||||
tokenizer = T5Tokenizer.from_pretrained(model_name)
|
||||
|
||||
# Set the device to use
|
||||
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
||||
model.to(device)
|
||||
|
||||
sample_chunks = ["You all just came off of your incredible Google Cloud next conference where you released a wide variety of functionality and features and new products across artisan television and also across the entire sort of cloud ecosystem . You want to just first by walking through , first start by walking through all the innovations that you sort of released and what you 're excited about when you come to Google Cloud ? Now our vision is super simple . If you look at what smartphones did for a consumer , you know they took a computer and internet browser , a communication device , and a camera , and made it so that it 's in everybody 's pocket , so it really brought computation to every person . We feel that , you know , our , what we 're trying to do is take all the technological innovation that Google 's doing , but make it super simple so that everyone can consume it . And so that includes our global data center footprint , all the new types of hardware and large-scale systems we work on , the software that we 're making available for people to do high-scale computation , tools for data processing , tools for cybersecurity , processing , tools for cyber security , tools for machine learning , but make it so simple that everyone can use it . And every step that we do to simplify things for people , we think adoption can grow . And so that 's a lot of what we 've done these last three , four years , and we made a number of announcements that next in machine learning and AI in particular , you know , we look at our work as four elements , how we take our large-scale compute systems that were building for AI and how we make that available to everybody . Second , what we 're doing with the software stacks and top of it , things like jacks and other things and how we 're making those available to everybody . Third is advances because different people have different levels of expertise . Some people say I need the hardware to build my own large language model or algorithm . Other people say , look , I really need to use a building block . You guys give me . So , 30s we 've done a lot with AutoML and we announce new capability for image , video , and translation to make it available to everybody . And then lastly , we 're also building completely packaged solutions for some areas and we announce some new stuff . ",
|
||||
" We 're joined next by Thomas Curian , CEO of Google Cloud , and Alexander Wang , CEO and founder of Scale AI . Thomas joined Google in November 2018 as the CEO of Google Cloud . Prior to Google , Thomas spent 22 years at Oracle , where most recently he was president of product development . Before that , Thomas worked at McKinsey as a business analyst and engagement manager . His nearly 30 years of experience have given him a deep knowledge of engineering enterprise relationships and leadership of large organizations . Thomas 's degrees include an MBA in administration and management from Stanford University , as an RJ Miller scholar and a BSEE in electrical engineering and computer science from Princeton University , where he graduated suma cum laude . Thomas serves as a member of the Stanford graduate School of Business Advisory Council and Princeton University School of Engineering Advisory Council . Please welcome to the stage , Thomas Curian and Alexander Wang . This is a super exciting conversation . Thanks for being here , Thomas ."]
|
||||
|
||||
|
||||
# Define the input text for summarization
|
||||
text = "Summarize the following text in 3 key points. text : " + sample_chunks[1]
|
||||
|
||||
# Tokenize the input text
|
||||
inputs = tokenizer.encode(text, return_tensors="pt").to(device)
|
||||
|
||||
# Generate the summary
|
||||
summary_ids = model.generate(inputs, max_length=1000, num_beams=4, early_stopping=True)
|
||||
|
||||
# Decode and print the summary
|
||||
summary = tokenizer.decode(summary_ids.squeeze(), skip_special_tokens=True)
|
||||
print("Summary:", summary)
|
||||
File diff suppressed because one or more lines are too long
@@ -1,44 +0,0 @@
|
||||
from gpt4all import GPT4All
|
||||
|
||||
model = GPT4All("/Users/gokulmohanarangan/Library/Application Support/nomic.ai/GPT4All/ggml-vicuna-13b-1.1-q4_2.bin")
|
||||
|
||||
import spacy
|
||||
|
||||
|
||||
def split_text_file(filename, token_count):
|
||||
nlp = spacy.load('en_core_web_md')
|
||||
|
||||
with open(filename, 'r') as file:
|
||||
text = file.read()
|
||||
|
||||
doc = nlp(text)
|
||||
total_tokens = len(doc)
|
||||
|
||||
parts = []
|
||||
start_index = 0
|
||||
|
||||
while start_index < total_tokens:
|
||||
end_index = start_index + token_count
|
||||
part_tokens = doc[start_index:end_index]
|
||||
part = ' '.join(token.text for token in part_tokens)
|
||||
parts.append(part)
|
||||
start_index = end_index
|
||||
|
||||
return parts
|
||||
|
||||
parts = split_text_file("transcript.txt", 1800)
|
||||
final_summary = []
|
||||
for part in parts:
|
||||
prompt = f"""
|
||||
### Human:
|
||||
Summarize the following text without missing any key points and action items.
|
||||
|
||||
{part}
|
||||
### Assistant:
|
||||
"""
|
||||
output = model.generate(prompt)
|
||||
final_summary.append(output)
|
||||
|
||||
|
||||
with open("sum.txt", "w") as sum:
|
||||
sum.write(" ".join(final_summary))
|
||||
@@ -1,8 +0,0 @@
|
||||
AGENDA: Most important things to look for in a start up
|
||||
TAM: Make sure the market is sufficiently large than once they win they can get rewarded
|
||||
Product market fit: Being in a good market with a product than can satisfy that market
|
||||
Unit economics: Profit for delivering all-in cost must be attractive (% or $ amount)
|
||||
LTV CAC: Life-time value (revenue contribution) vs cost to acquire customer must be healthy
|
||||
Churn: Fits into LTV, low churn leads to higher LTV and helps keep future CAC down
|
||||
Business: Must have sufficient barriers to entry to ward off copy-cats once established
|
||||
Founders: Must be religious about their product. Believe they will change the world against all odds.
|
||||
@@ -1,183 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
|
||||
# summarize https://www.youtube.com/watch?v=imzTxoEDH_g
|
||||
# summarize https://www.sprocket.org/video/cheesemaking.mp4 summary.txt
|
||||
# summarize podcast.mp3 summary.txt
|
||||
|
||||
import argparse
|
||||
import os
|
||||
import re
|
||||
import subprocess
|
||||
import tempfile
|
||||
from datetime import datetime
|
||||
from urllib.parse import urlparse
|
||||
|
||||
import jax.numpy as jnp
|
||||
import moviepy.editor
|
||||
import nltk
|
||||
import yt_dlp as youtube_dl
|
||||
from whisper_jax import FlaxWhisperPipline
|
||||
|
||||
from ...utils.file_utils import download_files, upload_files
|
||||
from ...utils.log_utils import LOGGER
|
||||
from ...utils.run_utils import CONFIG
|
||||
from ...utils.text_utils import post_process_transcription, summarize
|
||||
from ...utils.viz_utils import create_talk_diff_scatter_viz, create_wordcloud
|
||||
|
||||
nltk.download('punkt', quiet=True)
|
||||
nltk.download('stopwords', quiet=True)
|
||||
|
||||
WHISPER_MODEL_SIZE = CONFIG['WHISPER']["WHISPER_MODEL_SIZE"]
|
||||
NOW = datetime.now()
|
||||
|
||||
if not os.path.exists('../../artefacts'):
|
||||
os.makedirs('../../artefacts')
|
||||
|
||||
|
||||
def init_argparse() -> argparse.ArgumentParser:
|
||||
"""
|
||||
Parse the CLI arguments
|
||||
:return: parser object
|
||||
"""
|
||||
parser = argparse.ArgumentParser(
|
||||
usage="%(prog)s [OPTIONS] <LOCATION> <OUTPUT>",
|
||||
description="Creates a transcript of a video or audio file, then"
|
||||
" summarizes it using ChatGPT."
|
||||
)
|
||||
|
||||
parser.add_argument("-l", "--language",
|
||||
help="Language that the summary should be written in",
|
||||
type=str,
|
||||
default="english",
|
||||
choices=['english', 'spanish', 'french', 'german',
|
||||
'romanian'])
|
||||
parser.add_argument("location")
|
||||
return parser
|
||||
|
||||
|
||||
def main():
|
||||
parser = init_argparse()
|
||||
args = parser.parse_args()
|
||||
|
||||
# Parse the location string that was given to us, and figure out if it's a
|
||||
# local file (audio or video), a YouTube URL, or a URL referencing an
|
||||
# audio or video file.
|
||||
url = urlparse(args.location)
|
||||
|
||||
# S3 : Pull artefacts to S3 bucket ?
|
||||
|
||||
media_file = ""
|
||||
if url.scheme == 'http' or url.scheme == 'https':
|
||||
# Check if we're being asked to retreive a YouTube URL, which is
|
||||
# handled differently, as we'll use a secondary site to download
|
||||
# the video first.
|
||||
if re.search('youtube.com', url.netloc, re.IGNORECASE):
|
||||
# Download the lowest resolution YouTube video
|
||||
# (since we're just interested in the audio).
|
||||
# It will be saved to the current directory.
|
||||
LOGGER.info("Downloading YouTube video at url: " + args.location)
|
||||
|
||||
# Create options for the download
|
||||
ydl_opts = {
|
||||
'format': 'bestaudio/best',
|
||||
'postprocessors': [{
|
||||
'key': 'FFmpegExtractAudio',
|
||||
'preferredcodec': 'mp3',
|
||||
'preferredquality': '192',
|
||||
}],
|
||||
'outtmpl': './artefacts/audio', # Specify output file path and name
|
||||
}
|
||||
|
||||
# Download the audio
|
||||
with youtube_dl.YoutubeDL(ydl_opts) as ydl:
|
||||
ydl.download([args.location])
|
||||
media_file = "../artefacts/audio.mp3"
|
||||
|
||||
LOGGER.info("Saved downloaded YouTube video to: " + media_file)
|
||||
else:
|
||||
# XXX - Download file using urllib, check if file is
|
||||
# audio/video using python-magic
|
||||
LOGGER.info(f"Downloading file at url: {args.location}")
|
||||
LOGGER.info(" XXX - This method hasn't been implemented yet.")
|
||||
elif url.scheme == '':
|
||||
media_file = url.path
|
||||
# If file is not present locally, take it from S3 bucket
|
||||
if not os.path.exists(media_file):
|
||||
download_files([media_file])
|
||||
|
||||
if media_file.endswith(".m4a"):
|
||||
subprocess.run(["ffmpeg", "-i", media_file, f"./artefacts/{media_file}.mp4"])
|
||||
media_file = f"./artefacts/{media_file}.mp4"
|
||||
else:
|
||||
print("Unsupported URL scheme: " + url.scheme)
|
||||
quit()
|
||||
|
||||
# Handle video
|
||||
if not media_file.endswith(".mp3"):
|
||||
try:
|
||||
video = moviepy.editor.VideoFileClip(media_file)
|
||||
audio_filename = tempfile.NamedTemporaryFile(suffix=".mp3",
|
||||
delete=False).name
|
||||
video.audio.write_audiofile(audio_filename, logger=None)
|
||||
LOGGER.info(f"Extracting audio to: {audio_filename}")
|
||||
# Handle audio only file
|
||||
except Exception:
|
||||
audio = moviepy.editor.AudioFileClip(media_file)
|
||||
audio_filename = tempfile.NamedTemporaryFile(suffix=".mp3",
|
||||
delete=False).name
|
||||
audio.write_audiofile(audio_filename, logger=None)
|
||||
else:
|
||||
audio_filename = media_file
|
||||
|
||||
LOGGER.info("Finished extracting audio")
|
||||
LOGGER.info("Transcribing")
|
||||
# Convert the audio to text using the OpenAI Whisper model
|
||||
pipeline = FlaxWhisperPipline("openai/whisper-" + WHISPER_MODEL_SIZE,
|
||||
dtype=jnp.float16,
|
||||
batch_size=16)
|
||||
whisper_result = pipeline(audio_filename, return_timestamps=True)
|
||||
LOGGER.info("Finished transcribing file")
|
||||
|
||||
whisper_result = post_process_transcription(whisper_result)
|
||||
|
||||
transcript_text = ""
|
||||
for chunk in whisper_result["chunks"]:
|
||||
transcript_text += chunk["text"]
|
||||
|
||||
with open("./artefacts/transcript_" + NOW.strftime("%m-%d-%Y_%H:%M:%S") +
|
||||
".txt", "w") as transcript_file:
|
||||
transcript_file.write(transcript_text)
|
||||
|
||||
with open("./artefacts/transcript_with_timestamp_" +
|
||||
NOW.strftime("%m-%d-%Y_%H:%M:%S") + ".txt",
|
||||
"w") as transcript_file_timestamps:
|
||||
transcript_file_timestamps.write(str(whisper_result))
|
||||
|
||||
LOGGER.info("Creating word cloud")
|
||||
create_wordcloud(NOW)
|
||||
|
||||
LOGGER.info("Performing talk-diff and talk-diff visualization")
|
||||
create_talk_diff_scatter_viz(NOW)
|
||||
|
||||
# S3 : Push artefacts to S3 bucket
|
||||
prefix = "./artefacts/"
|
||||
suffix = NOW.strftime("%m-%d-%Y_%H:%M:%S")
|
||||
files_to_upload = [prefix + "transcript_" + suffix + ".txt",
|
||||
prefix + "transcript_with_timestamp_" + suffix + ".txt",
|
||||
prefix + "df_" + suffix + ".pkl",
|
||||
prefix + "wordcloud_" + suffix + ".png",
|
||||
prefix + "mappings_" + suffix + ".pkl",
|
||||
prefix + "scatter_" + suffix + ".html"]
|
||||
upload_files(files_to_upload)
|
||||
|
||||
summarize(transcript_text, NOW, False, False)
|
||||
|
||||
LOGGER.info("Summarization completed")
|
||||
|
||||
# Summarization takes a lot of time, so do this separately at the end
|
||||
files_to_upload = [prefix + "summary_" + suffix + ".txt"]
|
||||
upload_files(files_to_upload)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -1,151 +0,0 @@
|
||||
#!/usr/bin/env python3
|
||||
|
||||
import time
|
||||
import wave
|
||||
from datetime import datetime
|
||||
|
||||
import jax.numpy as jnp
|
||||
import pyaudio
|
||||
from pynput import keyboard
|
||||
from termcolor import colored
|
||||
from whisper_jax import FlaxWhisperPipline
|
||||
|
||||
from ...utils.file_utils import upload_files
|
||||
from ...utils.log_utils import LOGGER
|
||||
from ...utils.run_utils import CONFIG
|
||||
from ...utils.text_utils import post_process_transcription, summarize
|
||||
from ...utils.viz_utils import create_talk_diff_scatter_viz, create_wordcloud
|
||||
|
||||
WHISPER_MODEL_SIZE = CONFIG['WHISPER']["WHISPER_MODEL_SIZE"]
|
||||
|
||||
FRAMES_PER_BUFFER = 8000
|
||||
FORMAT = pyaudio.paInt16
|
||||
CHANNELS = 2
|
||||
RATE = 96000
|
||||
RECORD_SECONDS = 15
|
||||
NOW = datetime.now()
|
||||
|
||||
|
||||
def main():
|
||||
p = pyaudio.PyAudio()
|
||||
AUDIO_DEVICE_ID = -1
|
||||
for i in range(p.get_device_count()):
|
||||
if p.get_device_info_by_index(i)["name"] == \
|
||||
CONFIG["AUDIO"]["BLACKHOLE_INPUT_AGGREGATOR_DEVICE_NAME"]:
|
||||
AUDIO_DEVICE_ID = i
|
||||
audio_devices = p.get_device_info_by_index(AUDIO_DEVICE_ID)
|
||||
stream = p.open(
|
||||
format=FORMAT,
|
||||
channels=CHANNELS,
|
||||
rate=RATE,
|
||||
input=True,
|
||||
frames_per_buffer=FRAMES_PER_BUFFER,
|
||||
input_device_index=int(audio_devices['index'])
|
||||
)
|
||||
|
||||
pipeline = FlaxWhisperPipline("openai/whisper-" +
|
||||
CONFIG["WHISPER"]["WHISPER_REAL_TIME_MODEL_SIZE"],
|
||||
dtype=jnp.float16,
|
||||
batch_size=16)
|
||||
|
||||
transcription = ""
|
||||
|
||||
TEMP_AUDIO_FILE = "temp_audio.wav"
|
||||
global proceed
|
||||
proceed = True
|
||||
|
||||
def on_press(key):
|
||||
if key == keyboard.Key.esc:
|
||||
global proceed
|
||||
proceed = False
|
||||
|
||||
transcript_with_timestamp = {"text": "", "chunks": []}
|
||||
last_transcribed_time = 0.0
|
||||
|
||||
listener = keyboard.Listener(on_press=on_press)
|
||||
listener.start()
|
||||
print("Attempting real-time transcription.. Listening...")
|
||||
|
||||
try:
|
||||
while proceed:
|
||||
frames = []
|
||||
start_time = time.time()
|
||||
for i in range(0, int(RATE / FRAMES_PER_BUFFER * RECORD_SECONDS)):
|
||||
data = stream.read(FRAMES_PER_BUFFER,
|
||||
exception_on_overflow=False)
|
||||
frames.append(data)
|
||||
end_time = time.time()
|
||||
|
||||
wf = wave.open(TEMP_AUDIO_FILE, 'wb')
|
||||
wf.setnchannels(CHANNELS)
|
||||
wf.setsampwidth(p.get_sample_size(FORMAT))
|
||||
wf.setframerate(RATE)
|
||||
wf.writeframes(b''.join(frames))
|
||||
wf.close()
|
||||
|
||||
whisper_result = pipeline(TEMP_AUDIO_FILE, return_timestamps=True)
|
||||
timestamp = whisper_result["chunks"][0]["timestamp"]
|
||||
start = timestamp[0]
|
||||
end = timestamp[1]
|
||||
if end is None:
|
||||
end = start + 15.0
|
||||
duration = end - start
|
||||
item = {'timestamp': (last_transcribed_time,
|
||||
last_transcribed_time + duration),
|
||||
'text': whisper_result['text'],
|
||||
'stats': (str(end_time - start_time), str(duration))
|
||||
}
|
||||
last_transcribed_time = last_transcribed_time + duration
|
||||
transcript_with_timestamp["chunks"].append(item)
|
||||
transcription += whisper_result['text']
|
||||
|
||||
print(colored("<START>", "yellow"))
|
||||
print(colored(whisper_result['text'], 'green'))
|
||||
print(colored("<END> Recorded duration: " +
|
||||
str(end_time - start_time) +
|
||||
" | Transcribed duration: " +
|
||||
str(duration), "yellow"))
|
||||
|
||||
except Exception as exception:
|
||||
print(str(exception))
|
||||
finally:
|
||||
with open("real_time_transcript_" + NOW.strftime("%m-%d-%Y_%H:%M:%S")
|
||||
+ ".txt", "w", encoding="utf-8") as file:
|
||||
file.write(transcription)
|
||||
|
||||
with open("real_time_transcript_with_timestamp_" +
|
||||
NOW.strftime("%m-%d-%Y_%H:%M:%S") + ".txt", "w",
|
||||
encoding="utf-8") as file:
|
||||
transcript_with_timestamp["text"] = transcription
|
||||
file.write(str(transcript_with_timestamp))
|
||||
|
||||
transcript_with_timestamp = \
|
||||
post_process_transcription(transcript_with_timestamp)
|
||||
|
||||
LOGGER.info("Creating word cloud")
|
||||
create_wordcloud(NOW, True)
|
||||
|
||||
LOGGER.info("Performing talk-diff and talk-diff visualization")
|
||||
create_talk_diff_scatter_viz(NOW, True)
|
||||
|
||||
# S3 : Push artefacts to S3 bucket
|
||||
suffix = NOW.strftime("%m-%d-%Y_%H:%M:%S")
|
||||
files_to_upload = ["real_time_transcript_" + suffix + ".txt",
|
||||
"real_time_transcript_with_timestamp_" + suffix + ".txt",
|
||||
"real_time_df_" + suffix + ".pkl",
|
||||
"real_time_wordcloud_" + suffix + ".png",
|
||||
"real_time_mappings_" + suffix + ".pkl",
|
||||
"real_time_scatter_" + suffix + ".html"]
|
||||
upload_files(files_to_upload)
|
||||
|
||||
summarize(transcript_with_timestamp["text"], NOW, True, True)
|
||||
|
||||
LOGGER.info("Summarization completed")
|
||||
|
||||
# Summarization takes a lot of time, so do this separately at the end
|
||||
files_to_upload = ["real_time_summary_" + suffix + ".txt"]
|
||||
upload_files(files_to_upload)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
Reference in New Issue
Block a user