Je n’ai rien appris beaucoup de nouveaux mots aujourd’hui, mais j’ai appris quelque choses de grammaire. Je sais maintenant que l’ordre de plusieurs adjectifs avant le nom est très complexe. Il n’y pas des règles précises quand il y a plusieurs adjectifs. J’ai aussi appris comment utiliser un adverbe dans le passé composé.
| French | English | Notes |
|---|---|---|
| Je <<vais bientôt terminer>> ce livre | I am going to finish this book soon. | Notice the adverb placement |
Quand j’étais petite, j’ai écouté mes cousins parlent en francais. Depuis, j’ai adoré le langue est j’ai pris français à l’université. j’ai regardé aussi French In Action, une serie de PBS. J’ai visité une fois en 1999 à Paris.
J’adore la langue et la nourriture, et je veux visiter encore est habiter là. Avant, je veux apprendre français si je peux le parler courament. Je voudrai écrire et parle le français plus mieux qu’aujourd’hui.
Si maintenant, j’étudie le français tous les jours. J’utilise Duolingo et mes notes à l’apprendre. Je le recherche aussi les mots et grammaire avec sites web comme Lawless French, Reverso et French Language Stack Exchange.
Ici, je veux à partager mes notes de la langue parce qu’ils contiennent mes questions sur les leçons. Depuis, je peux les chercher quand je veux.
Codelab 2/2 from day one. The code lab is here in Kaggle.
"""
Prompting
Google Gen AI 5-Day Intensive Course
Host: Kaggle
Day: 1
Codelab: https://www.kaggle.com/code/reniseblack/day-1-evaluation-and-structured-output/edit
"""
import enum
import io
import os
from pprint import pprint
import typing_extensions as typing
from google import genai
from google.api_core import retry
from google.genai import types
from IPython.display import Markdown, clear_output, display
is_retriable = lambda e: ( # noqa: E731
isinstance(e, genai.errors.APIError) and e.code in {429, 503})
genai.models.Models.generate_content = retry.Retry(
predicate=is_retriable)(genai.models.Models.generate_content)
client = genai.Client(api_key=os.environ["GOOGLE_API_KEY"])
# Create and send a single-turn prompt
MODEL = "gemini-2.0-flash"
prompt = "Explain AI to me like I'm a kid."
response = client.models.generate_content(
model=MODEL,
contents=prompt
)
# Get the response text then render as HTML (optional)
response_text = response.text
print("Model response: \n\n", Markdown(response.text))
section_break = "----------------------------"
print(section_break, "\n\nStarting a multi-turn chat ...\n\n")
# Start a multi-turn chat
chat_message = "Hello! My name is Zlork."
chat = client.chats.create(model=MODEL, history=[])
response = chat.send_message(chat_message)
print(response.text)
chat_message2 = "Can you tell me something interesting about dinosaurs?"
response = chat.send_message(chat_message2)
print(response.text)
chat_message3 = "Do you remember my name?"
response = chat.send_message(chat_message3)
print(response.text)
# See the list of available models
print("\n\nGetting a list of available models ... \n\n")
for model in client.models.list():
print(model.name)
print("\n\nGetting a list of gemini-2.0-flash models ... \n\n")
for model in client.models.list():
if model.name == 'models/gemini-2.0-flash':
pprint(model.to_json_dict())
break
# Set token limit to restrict output length
print("\n\nSetting a token limit ... \n\n")
short_config = types.GenerateContentConfig(max_output_tokens=200)
PROMPT_2 = ('Write a 1000 word essay on the importance of olives in modern '
'society.')
response = client.models.generate_content(
model=MODEL,
config=short_config,
contents=PROMPT_2
)
print(response.text)
print("\n\n")
PROMPT_3 = "Write a short poem on the importance of olives in modern society."
# config is where you specify top_k, temperature, etc
response = client.models.generate_content(
model=MODEL,
config=short_config,
contents=PROMPT_3
)
print(response.text)
print("\n\n")
short_config = types.GenerateContentConfig(max_output_tokens=100)
response = client.models.generate_content(
model=MODEL,
config=short_config,
contents='Write a 150 opinion about the importance of pine trees.')
print(response.text)
print("\n\n")
print("\n\nSetting a temperature limit ... \n\n")
# Set a temperature that helps manage the randomness
high_temp_config = types.GenerateContentConfig(temperature=2.0)
PROMPT_4 = "Pick a random colour... (respond in a single word)"
for _ in range (5):
response = client.models.generate_content(
model=MODEL,
config=high_temp_config,
contents = PROMPT_4
)
if response.text:
print(response.text, "-" * 25)
print("\n\n")
print("\n\nSetting a low temperature ... \n\n")
# Setting a low temperature
low_temp_config = types.GenerateContentConfig(temperature=0.0)
for _ in range(5):
response = client.models.generate_content(
model=MODEL,
config=low_temp_config,
contents = PROMPT_4
)
if response.text:
print(response.text, "-" * 25)
print("\n\n")
print("\n\nSetting top-p ... \n\n")
# Setting top-p that also controls output diversity
# It will determine when to stop selecting tokens that are least probable
TEMPERATURE_2 = 1.0
TOP_P1 = 0.05
model_config = types.GenerateContentConfig(
# These are default gemini-2.0-flash values
temperature=TEMPERATURE_2,
top_p=TOP_P1
)
STORY_PROMPT = ("You are a creative writer. Write a short story about a cat"\
"who goes on an adventure.")
response = client.models.generate_content(
model=MODEL,
config=model_config,
contents=STORY_PROMPT
)
print(response.text)
print("\n\n")
print("\n\nSetting top-p ... \n\n")
# Create a zero-shot prompt
TEMPERATURE_2 = 0.1
TOP_P2 = 1
MAX_OUTPUT_TOKENS = 5
model_config = types.GenerateContentConfig(
temperature=TEMPERATURE_2,
top_p=TOP_P2,
max_output_tokens=MAX_OUTPUT_TOKENS,
)
ZERO_SHOT_PROMPT = """Classify movie reviews as POSITIVE, NEUTRAL or NEGATIVE.
Review: "Her" is a disturbing study revealing the direction
humanity is headed if AI is allowed to keep evolving,
unchecked. I wish there were more movies like this masterpiece.
Sentiment: """
response = client.models.generate_content(
model=MODEL,
config=model_config,
contents=ZERO_SHOT_PROMPT)
print(response.text)
print("\n\n")
print("\n\nSetting top-p ... \n\n")
# Setting enum mode
class Sentiment(enum.Enum):
POSITIVE = "right on"
NEUTRAL = "meh"
NEGATIVE = "sucks lemons"
ENUM_TYPE = "text/x.enum"
response = client.models.generate_content(
model=MODEL,
config=types.GenerateContentConfig(
response_mime_type=ENUM_TYPE,
response_schema=Sentiment
),
contents=ZERO_SHOT_PROMPT)
print(response.text)
print("\n\n")
print("\n\nSetting top-p ... \n\n")
# Get the Sentiment returned as a Python object
enum_response = response.parsed
print(enum_response)
print(type(enum_response))
print("\n\n")
print("\n\nSetting top-p ... \n\n")
# Send one-shot and few-shot prompts
FEW_SHOT_PROMPT = """Parse a customer's pizza order into valid JSON:
EXAMPLE:
I want a small pizza with cheese, tomato sauce, and pepperoni.
JSON Response:
```
{
"size": "small",
"type": "normal",
"ingredients": ["cheese", "tomato sauce", "pepperoni"]
}
```
EXAMPLE:
Can I get a large pizza with tomato sauce, basil and mozzarella
JSON Response:
```
{
"size": "large",
"type": "normal",
"ingredients": ["tomato sauce", "basil", "mozzarella"]
}
```
ORDER:
"""
CUSTOMER_ORDER = "Give me a large with cheese & pineapple"
MAX_OUTPUT_TOKENS = 250
response = client.models.generate_content(
model='gemini-2.0-flash',
config=types.GenerateContentConfig(
temperature=TEMPERATURE_2,
top_p=TOP_P2,
max_output_tokens=MAX_OUTPUT_TOKENS,
),
contents=[FEW_SHOT_PROMPT, CUSTOMER_ORDER])
print(response.text)
print("\n\n")
print("\n\nSet to JSON response only ... \n\n")
# Set output to JSON format
class PizzaOrder(typing.TypedDict):
size: str
ingredients: list[str]
type: str
ORDER_1 = "Can I have a large dessert pizza with apple and chocolate"
response = client.models.generate_content(
model=MODEL,
config=types.GenerateContentConfig(
temperature=TEMPERATURE_2,
response_mime_type="application/json",
response_schema=PizzaOrder
),
contents=ORDER_1
)
print(response.text)
print("\n\n")
print("\n\nUse Chain of Thought (CoT) ... \n\n")
# Use Chain of Thought (CoT) to show reasoning to check hallucinations
CHAIN_OF_THOUGHT_PROMPT = """When I was 4 years old,
my partner was 3 times my age. Now, I am 20 years old.
How old is my partner? Return the answer directly."""
response = client.models.generate_content(
model=MODEL,
contents=CHAIN_OF_THOUGHT_PROMPT
)
print(response.text)
print("\n\n")
print("\n\nUse Chain of Thought (CoT) with step-by-step ... \n\n")
# Use Chain of Thought (CoT) to show reasoning to check hallucinations
CHAIN_OF_THOUGHT_PROMPT_2 = """When I was 4 years old, my partner was 3 times
my age. Now, I am 20 years old. How old is my partner?
Let's think step by step."""
response = client.models.generate_content(
model=MODEL,
contents=CHAIN_OF_THOUGHT_PROMPT_2
)
Markdown(response.text)
print("\n\n")
print("\n\nReAct prompt examples ... \n\n")
# Examples of ReAct prompts
model_instructions = """
Solve a question answering task with interleaving Thought, Action, Observation
steps. Thought can reason about the current situation,
Observation is understanding relevant information from an Action's output and
Action can be one of three types:
(1) <search>entity</search>, which searches the exact entity on Wikipedia and
returns the first paragraph if it exists. If not, it will return some
similar entities to search and you can try to search the information from
those topics.
(2) <lookup>keyword</lookup>, which returns the next sentence containing
keyword in the current context. This only does exact matches, so keep your
searches short.
(3) <finish>answer</finish>, which returns the answer and finishes the task.
"""
example1 = """Question
Musician and satirist Allie Goertz wrote a song about the "The Simpsons" ch\
aracter Milhouse, who Matt Groening named after who?
Thought 1
The question simplifies to "The Simpsons" character Milhouse is named after \
who. I only need to search Milhouse and find who it is named after.
Action 1
<search>Milhouse</search>
Observation 1
Milhouse Mussolini Van Houten is a recurring character in the Fox animated \
television series The Simpsons voiced by Pamela Hayden and
created by Matt Groening.
Thought 2
The paragraph does not tell who Milhouse is named after, maybe I can look u\
p "named after".
Action 2
<lookup>named after</lookup>
Observation 2
Milhouse was named after U.S. president Richard Nixon, whose middle name wa\
s Milhous.
Thought 3
Milhouse was named after U.S. president Richard Nixon, so the answer is Ric\
hard Nixon.
Action 3
<finish>Richard Nixon</finish>
"""
example2 = """Question
What is the elevation range for the area that the eastern sector of the
Colorado orogeny extends into?
Thought 1
I need to search Colorado orogeny, find the area that the eastern sector
of the Colorado orogeny extends into, then find the elevation range
of the area.
Action 1
<search>Colorado orogeny</search>
Observation 1
The Colorado orogeny was an episode of mountain building (an orogeny) in Colora
do and surrounding areas.
Thought 2
It does not mention the eastern sector. So I need to look up eastern sector.
Action 2
<lookup>eastern sector</lookup>
Observation 2
The eastern sector extends into the High Plains and is called the Central Plain
s orogeny.
Thought 3
The eastern sector of Colorado orogeny extends into the High Plains. So I need
to search High Plains and find its elevation range.
Action 3
<search>High Plains</search>
Observation 3
High Plains refers to one of two distinct land regions
Thought 4
I need to instead search High Plains (United States).
Action 4
<search>High Plains (United States)</search>
Observation 4
The High Plains are a subregion of the Great Plains. From east to west, the
High Plains rise in elevation from around 1,800 to 7,000 ft (550 to 2,130m).
Thought 5
High Plains rise in elevation from around 1,800 to 7,000 ft, so the answer is 1
,800 to 7,000 ft.
Action 5
<finish>1,800 to 7,000 ft</finish>
"""
question = """Question
Who was the youngest author listed on the transformers NLP paper?
"""
# You will perform the Action; so generate up to, but not including,
# the Observation.
react_config = types.GenerateContentConfig(
stop_sequences=["\nObservation"],
system_instruction=model_instructions + example1 + example2,
)
# Create a chat that has the model instructions and examples pre-seeded.
react_chat = client.chats.create(
model=MODEL,
config=react_config,
)
response = react_chat.send_message(question)
print(response.text)
print("\n\n")
print("\n\nAdding an observation with author names ... \n\n")
# The model won't find the authors, so supply them with a followup
observation = """Observation 1
[1706.03762] Attention Is All You Need
Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan
N. Gomez, Lukasz Kaiser, Illia Polosukhin
We propose a new simple network architecture, the Transformer, based solely on
attention mechanisms, dispensing with recurrence and convolutions entirely.
"""
response = react_chat.send_message(observation)
print(response.text)
print("\n\n")
print("\n\nUse Gemini 2.0 Flash Thinking Mode ... \n\n")
# Examples of ReAct prompts
THINKING_MODE_MODEL = "gemini-2.0-flash-thinking-exp"
THINKING_MODE_PROMPT = ("Who was the youngest author listed on the "
"transformers NLP paper?")
response = client.models.generate_content_stream(
model=THINKING_MODE_MODEL,
contents=THINKING_MODE_PROMPT
)
buffer = io.StringIO()
for chunk in response:
buffer.write(str(chunk.text))
# Display the response as it is streamed
print(chunk.text, end='')
# Render the response as formatted markdown
clear_output()
Markdown(buffer.getvalue())
print("\n\n")
print("\n\nUse Gemini 2.0 to generate code ... \n\n")
# Generate code wth Gemini
# The Gemini models love to talk, so it helps to specify they stick to the code
# if that is all that you want.
CODE_PROMPT = """Write a Python function to calculate the factorial of a number.
No explanation, provide only the code."""
response = client.models.generate_content(
model=MODEL,
config=types.GenerateContentConfig(
temperature=1,
top_p=1,
max_output_tokens=1024,
),
contents=CODE_PROMPT)
Markdown(response.text)
print("\n\n")
print("\n\nUse Gemini 2.0 to execute code ... \n\n")
# Execute code wth Gemini
config = types.GenerateContentConfig(
tools=[types.Tool(code_execution=types.ToolCodeExecution())],
)
code_exec_prompt = """
Generate the first 14 odd prime numbers, then calculate their sum.
"""
response = client.models.generate_content(
model=MODEL,
config=config,
contents=code_exec_prompt)
for part in response.candidates[0].content.parts:
pprint(part.to_json_dict())
print("-----")
print("\n\n")
print("\n\nSee more parts of the response ... \n\n")
# See executable_code and code_execution_result
for part in response.candidates[0].content.parts:
if part.text:
display(Markdown(part.text))
elif part.executable_code:
display(Markdown(f'```python\n{part.executable_code.code}\n```'))
elif part.code_execution_result:
if part.code_execution_result.outcome != 'OUTCOME_OK':
display(Markdown(f'## Status {part.code_execution_result.outcome}'))
display(Markdown(f'```\n{part.code_execution_result.output}\n```'))
print("\n\n")
print("\n\nExplaining code ... \n\n")
# Get an explanation for code
# file_contents = !curl https://raw.githubusercontent.com/magicmonty/bash-git-prompt/refs/heads/master/gitprompt.sh
explain_prompt = f"""
Please explain what this file does at a very high level. What is it,
and why would I use it?
```
{file_contents}
```
"""
response = client.models.generate_content(
model='gemini-2.0-flash',
contents=explain_prompt)
Markdown(response.text)Codelab 1/2 from day one. The code lab is here in Kaggle.
"""
Evaluation and Structured Output
Google Gen AI 5-Day Intensive Course
Host: Kaggle
Day: 1
Kaggle: https://www.kaggle.com/code/markishere/day-1-evaluation-and-structured-output
"""
import enum
import os
from google import genai
from google.api_core import retry
from google.genai import types
from IPython.display import Markdown, display
client = genai.Client(api_key=os.environ["GOOGLE_API_KEY"])
# Automated retry
is_retriable = lambda e: (
isinstance(e, genai.errors.APIError) and e.code in {429, 503}
)
genai.models.Models.generate_content = retry.Retry(predicate=is_retriable)(
genai.models.Models.generate_content
)
# if not hasattr(genai.models.Models.generate_content, '__wrapped__'):
# genai.models.Models.generate_content = retry.Retry(
# predicate=is_retriable)(genai.models.Models.generate_content)
# Evaluation
# Understand model performance
# Get the file locally first
# !wget -nv -O gemini.pdf https://storage.googleapis.com/cloud-samples-data/generative-ai/pdf/2403.05530.pdf
document_file = client.files.upload(file="/Users/renise/Documents/Python/gen_ai/day_one/gemini.pdf")
print("\n")
print(document_file)
print("\n")
print("\nSummarize a document\n")
# Summarize a document
def summarize_doc(request: str) -> str:
"""Execute the request on the uploaded document."""
# Set the temperature low to stabilize the output.
config = types.GenerateContentConfig(temperature=0.0)
response = client.models.generate_content(
model="gemini-2.0-flash",
config=config,
contents=[request, document_file],
)
return response.text
request = "Tell me about the training process used here."
summary = summarize_doc(request)
# display(Markdown(summary + "\n-----"))
print("\n\n")
# Define an evaluator
SUMMARY_PROMPT = """\
# Instruction
You are an expert evaluator. Your task is to evaluate the quality of the responses generated by AI models.
We will provide you with the user input and an AI-generated responses.
You should first read the user input carefully for analyzing the task, and then evaluate the quality of the responses based on the Criteria provided in the Evaluation section below.
You will assign the response a rating following the Rating Rubric and Evaluation Steps. Give step-by-step explanations for your rating, and only choose ratings from the Rating Rubric.
# Evaluation
## Metric Definition
You will be assessing summarization quality, which measures the overall ability to summarize text. Pay special attention to length constraints, such as in X words or in Y sentences. The instruction for performing a summarization task and the context to be summarized are provided in the user prompt. The response should be shorter than the text in the context. The response should not contain information that is not present in the context.
## Criteria
Instruction following: The response demonstrates a clear understanding of the summarization task instructions, satisfying all of the instruction's requirements.
Groundedness: The response contains information included only in the context. The response does not reference any outside information.
Conciseness: The response summarizes the relevant details in the original text without a significant loss in key information without being too verbose or terse.
Fluency: The response is well-organized and easy to read.
## Rating Rubric
5: (Very good). The summary follows instructions, is grounded, is concise, and fluent.
4: (Good). The summary follows instructions, is grounded, concise, and fluent.
3: (Ok). The summary mostly follows instructions, is grounded, but is not very concise and is not fluent.
2: (Bad). The summary is grounded, but does not follow the instructions.
1: (Very bad). The summary is not grounded.
## Evaluation Steps
STEP 1: Assess the response in aspects of instruction following, groundedness, conciseness, and verbosity according to the criteria.
STEP 2: Score based on the rubric.
# User Inputs and AI-generated Response
## User Inputs
### Prompt
{prompt}
## AI-generated Response
{response}
"""
# Define a structured enum class to capture the result.
class SummaryRating(enum.Enum):
VERY_GOOD = 5
GOOD = 4
OK = 3
BAD = 2
VERY_BAD = 1
def eval_summary(prompt, ai_response):
"""Evaluate the generated summary against the prompt."""
chat = client.chats.create(model="gemini-2.0-flash")
# Generate the full text response
response = chat.send_message(
message=SUMMARY_PROMPT.format(prompt=prompt, response=ai_response)
)
verbose_eval = response.text
# Coerce into desired structure
structured_output_config = types.GenerateContentConfig(
response_mime_type="text/x.enum",
response_schema=SummaryRating
)
response = chat.send_message(
message="Convert the final score.",
config=structured_output_config
)
structured_eval = response.parsed
return verbose_eval, structured_eval
text_eval, struct_eval = eval_summary(
prompt=[request, document_file],
ai_response=summary
)
Markdown(text_eval)
# Play with the summary prompt
new_prompt = "Explain like I'm 5 the training process"
# Try:
# ELI5 the training process
# Summarise the needle/haystack evaluation technique in 1 line
# Describe the model architecture to someone with a civil engineering degree
# What is the best LLM?
if not new_prompt:
raise ValueError("Try setting a new summarization prompt.")
def run_and_eval_summary(prompt):
"""Generate and evaluate the summary using the new prompt."""
summary = summarize_doc(new_prompt)
display(Markdown(summary + "\n-----"))
text, struct = eval_summary([new_prompt, document_file], summary)
display(Markdown(text + "\n-----"))
print(struct)
run_and_eval_summary(new_prompt)Les leçons de français était un peu difficile aujourd’hui parce que je n’ai concentré pas bien. Encore, j’ai fait asses bien. J’ai appris de passé composé avec les verbes voir, lire, terminé, et emprunté. J’ai appris aussi comment il faut dire “il y a ? ans” quand je parle du passé. Par exemple, il faut dire “Je l’ai vu il y a deux ans” pour quelque chose tu as vois deux ans du passé.
| French | English | Notes |
|---|---|---|
| Il <<en>> mange beaucoup | He eats a lot of <<it>> | |
| goût | taste | |
| Tu l’a lu? | ||
| Tu veux <<encore voir>> ce flim? | ||
| Je vais <<emprunter>> ce livre à la bibliothèque | I am going to <<borrow>> this book at the library | |
| Il veut être <<auteur>> | He wants to be <<an author>> | auteur (masc); |
| Ils l’ont <<vraiment adoré>> | They <<really loved>> it. | Notice “vraiment” is before the past tense of the action verb |
| Il a vraiment aimé <<le dernier livre de cet auteur>> | He really liked <<this author’s last book>> | Possessive nouns always come after what it’s possessing |
| J’ai <<déjà lu>> ce livre. Je ne veux pas encore le lire. | I <<already read>> this book. I don’t want to read it again. | |
| Il l’a déjà lu | He has already seen it | |
| Je <<n’ai pas terminé>> ce livre | I <<didn’t finished>> this book | |
| Je <<l’ai emprunté à>> Marc le mois dernier | I <<borrowed it from>> Marc last month | To borrow from is expressed with “à” instead of “de” |
| J’<<ai presque terminé>> ce livre | I <<almost finished>> this book | |
| J’<<ai presque tout lu>>. | I’ve <<almost read everything>> | |
| Tu as vu <<la fin du film>>? | Have you seen <<the end of the film >>? | |
| J’ai trente ans. <<Il y a dix ans>>, j’avais vingt ans | I am thirty years old. <<Ten years ago>>, I was 20 years old. | |
| Je vais <<commander>> ce livre <<sur Internet>> aujourd’hui. | I am going to <<order>> this book <<online>> today. |
C’est intéressant que les mots pour “teaspoon” en français sont “cuillère à café” mais les mots pour “tablespoon” sont “cuillère à soupe”.
| French | English | Notes |
|---|---|---|
| la cuillère | the (tea)spoon | |
| Il mange des saucisses | He is eating (some) sausage | |
| Je mets <<une cuillère à cafe>> de sucre>> | I put in <<a teaspoon>> of sugar | Notice how to say “teaspoon” |
| Il prépare <<de la purée>> de légumes | He is making (some) puréed vegetables | Notice the use of “de la” (fem) for the unknown quantity of puréed vegetables |
| J’ai besoin d’<<une cuillère à soupe>> | I need a tablespoon | Notice there difference between ‘Tablespoon’ here and ‘teaspoon’ above; also the use of à when specifying the type of spoon |
| la poire | the pear | |
| J’aime manger <<du pain grillé>> chaud avec du beurre et de la confiture | I like eating warm <<(some) toast>> with butter and jam | |
| Mon sandwich est simple: <<deux tranches de pain>> avec du jambon | My sandwich is simple: two slices of bread with (some) ham | |
| le citron | the lemon | |
| le champignon | the mushroom | |
| l’oignon | the onion | |
| la carotte | the carrot | |
| Je n’en vas pas | I don’t want any | |
| Je mets cent grammes de sucre | I <<am adding>> 100 grams of sugar | |
| J’en ai besoin <<pour mon plat>> ce soir | I need some <<for my dinner>> tonight | |
| Quelqu’un <<a volé>> mon déjeuner | Someone <<stole>> my lunch | |
| <<On me l’a volé>> tous les jours cette semaine | <<Someone has stolen it from me>> every day this week | |
| Tu <<veux dire>> au poulet | You <<mean>> chicken | |
| Tu <<me l’a déjà dit>>, tu <<ne t’en souviens pas>>? | You <<already told me that>>, <<don’t you remember>>? | |
| Il achète <<de la brioche>>. Il <<en>> achète. | He buys <<(some) brioche>>. He buys <<some>> | A good example for how to use “en” to replace “de la <noun>” phrases |
| Il faut ajouter du sel | You must add salt | |
| Tu veux des saucisses? | Do you want some sausage? | Notice “des” use for singular and plural sausage |
| Voulez-vous <<un peu de>> crème? | Do you want <<a little cream> | When another quantity is given, “de la/du/des” isn’t used |
Aujourd’hui, j’ai apris beaucoup des mots et grammaires. Je comprends comment à dire “il y a du vent dehors”, “une araignée”, “le toit”, “à l’étage”, “la pastèque”, y autres. J’ai passé du temps rechercher comment utiliser les prépositions “à”, “par” et “dans”. Je le sais plus mieux qu’avant.
| French | English | Notes |
|---|---|---|
| Elle est <<une femme de ménage>> | She is a <<cleaning lady>> | “Ménager” means “to clean”. |
| Il y a <<une araignée>> <<au plafond>> | There’s <<a spider>> <<on the ceiling>> | How to say both “spider” and “on the ceiling”. Notice “on the” is “au/à la” in french. |
| le toit | the roof (of a building) | |
| Ils sont venue <<à vélo>>? | They came <<by bike>>. | The preposition “à” implies “riding on” something |
| Il est venu hier | He came yesterday. | |
| L’oiseau est entré <<par la fenêtre>> | The bird entered <<through the window>> | Use “par” when going through something or indicating direction |
| Elles sont entrées <<dans le musée>>. | They entered <<into the museum>> | |
| <<Hier soir>>, il est parti à neuf heures. | <<Last night>>, she left at nine o’clock | |
| Il <<est descendu>> maintenant. | He <<came>> down now | Descendre can also mean to go down. |
| Ils <<sont montés>> <<à l’étage>>. | They <<went>> <<upstairs>>. | How to say “upstairs” |
| <<L’homme de ménage>> est venu nettoyer la maison | <<The cleaning man>> came to clean the house | |
| Il ne faut pas monter <<sur le toit>>. | One must not climb up <<on the roof>> | “Monter” means to “mount”, “climb”, “ride”, “go up”, “rise” depending on the context |
| Les araignées sont montées <<sur le mur>> | The spiders climbed up <<(on) the wall>> | To make it explicit that an insect climbs “on” the wall versus “up” the wall. |
| <<Il y a du vent>> dehors. | It’s windy outside. | This is a good phrase to know. |
| Il pleut. | It’s raining. | Another good phrase to know. |
| une araignée <<effrayante>>. | A <<scary>> spider. | Effrayant(e): scary, frightening, creepy, spooky, etc |
| Elle a <<une double de clé>> | She has <<a spare key>> | So “une double de” something means a duplicate of. |
| Tu <<peux en trouver>> au marché | You <<can find it>> at the market | The phrase “Tu peux en trouver” is common and good to know. |
| Non, je <<n’en mange pas>> | I <<don’t eat any>> | How to say “don’t (something) any”. |
| Combien de <<tranches de jambon>> voulez-vous? | How many <<slices of ham>> do you want? | |
| Sarah, tu manges <<de la viande>>? Oui, j’<<en>> mange parfois | Sarah, do you eat <<(some) meat>>? Yes, I eat <<some>> sometimes | Only with noun forms prefixed by “de” like “de la viande” |
| <<La pastèque>> est <<un gros>> fruit. | <<Watermelon>> is <<a big>> fruit | The word for “watermelon” |
| Tu veux de la pastèque? J’<<en>> veux. | Do you want <<some>> watermelon. I want some. | Another example with “en” |
| J’ai acheté <<de l’huile à l’ail>> | I bought <<some garlic oil>>. | How to combine “l’huile” and “l’ail” which is challenging to pronounce together. |
| Il déteste les légumes et il <<n’en mange jamais>> | He hates vegetables and he <<never eats any>>. | How to say “never eats any” or something similar |
J’ai appris les mots et les verbes très intéressants comme <<grossir>>, <<maigrir>> et <<ours>>. J’ai appris aussi comment à dire les matériaux de vêtements et bijoux avec <<en>>.
| French | English | Notes |
|---|---|---|
| J’ai <<plusieurs>> robes, quatre ou cinq. | I have several dresses, four or five | |
| C’est <<en promotion>>? | It’s <<on sale>>? | |
| J’ai <<maigri>> | I <<lost weight>> | maigrir (v); passé composé: maigri |
| Où est <<mon short>>? | Where are <<my shorts>>? | “shorts” in French doesn’t have an “s” at the end |
| Mon chat a <<grossi>> | My cat gained weight | grossir (v); passé composé: grossi |
| J’ai choisi <<une paire>> de <<baskets>> | I chose <<a pair>> of <<sneakers>> | <<paire>> est fém; <<baskets>> est avec <s> |
| J’ai une paire de baskets <<toutes neuves>> | I have a brand-new pair of sneakers | feminine of neuf: neuves and how to say “brand new” |
| Mes sacs sont <<bleu clair>> | My bags are <<light bleu>> | clair (masc); claire(fém) |
| Mes bottes sont <<bleu foncé>> | My boots are <<dark blue>> | foncé (masc. singular); foncée (fém. singular) |
| Ces chaussures sont <<parfaites>> | These shoes are <<perfect>> | parfait (masc. singular); parfaite (fém. singular) |
| Ces chaussures sont très <<confortables>> | These shoes are very <<comfortable>> | confortables (plural) |
| Ces chaussures et ce sac sont <<en cuir de vache>> | These shoes and this bag are leather | (le) cuir |
| Je porte <<cette bague>> parce que je suis mariée | I wear this ring because I am married | (une) bague |
| C’est <<une bague en argent>> | This is <<a silver ring>> | |
| Son pantalons est <<en coton>> | His pants are cotton | To specify clothing material prefix with “en” |
| J’ai un bracelet <<en or>> | I have a gold bracelet | Specify jewelry material prefixed with “en” |
| Elle a rempli l’armoire | She filled the closet | |
| Tu as <<trop de paires>> de chaussures | You have <<too many pairs>> of shoes | |
| J’adore ta jupe. Elle est <<neuve ou vieille>>? | I love your skirt. Is it <<new or old>>? | Féminin de “neuf” et “vieux” |
| Une tente génial! | A great tent | |
| Ils sont dans une tente dans <<la forêt>> | The are in a tent in <<the forest>> | |
| Il faut nuit | It’s night | How to use this common phrase |
| Tu aimes <<être là>>? | Do you like <<being here>>? | The use of “là” instead of “ici” |
| Je n’ai pas faim du tout | I am not hungry at all | Common saying |
| Oui, c’est <<presque>> parfait | Yes, it’s <<nearly>> perfect | |
| Qu’est-ce qui ne vas pas? | What’s wrong? | Common saying |
| C’est <<juste qu’il y a un ours>> dehors! | It’s <<just that there is a bear>> outside! | The phrase “just that there is” is good to know |
| Cette couleur est parfaite pour le monde | This color is perfect for everyone. | |
| Il va <<grandir>> trop vite | He is going <<to grow up>> too fast | |
| Tu n’as pas choisi <<la bonne taille>> | You didn’t choose <<the right size>> | |
| Ses pieds <<ne vont plus grandir>> | His feet <<won’t grow anymore>> | |
| Vous avez grandi! | You’ve grown! | |
| Les chapeaux <<de ce magasin>> sont parfaits. | The hats in this store are perfect. |
CEFR: 36
Aujourd’hui, j’ai appris beaucoup de choses. J’ai appris la ville de Paris. J’ai appris aussi <<voudrions>>, <<le plus proche>>, <<faire pour aller là-bas>>, <<rompre>>, <<déménager>>, <<vendre de commencer>>, <<vous vous retournez>>, <<grandir>>, <<la bonne taille>>, et autres en bas.
| French | English | Notes |
|---|---|---|
| Où est le café <<le plus proche?>> | Where is <<the closest>> cafe | |
| Excusez-moi, je suis <<perdue>> | Excuse me, I am <<lost>> | |
| Nous <<voudrions>> aller à la gare | We <<would like>> to go to the train station | |
| Comment est-ce que je <<fais pour aller là-bas>>? | How do I get over there? | This is a common saying I should know |
| Il y a <<tellement de>> voitures | There are <<so many>> cars | The use of “de” for singular and plural |
| Qui a <<le plan>>? | Who has <<the map>> | |
| Berlin est <<une capitale>> | Berlin is a <<capital city>> | Notice the gender |
| J’habite dans <<le centre-ville>> | I live in <<the city center>> | |
| Si <<tu te retournes>> je vais prendre une photo>> | If <<you turn (yourself) around>> I am going to take a photo | |
| Il ne faut pas traverser ici | One must not cross here | Something important to understand |
| <<Tu romps>> avec moi? | Are you breaking up with me | rompre (v): to break up |
| Ca <<ne vas pas être>> long | It won’t be long | |
| Mais, <<nous devons rompre>> | But, we have to break up | |
| Je dois déménager pour le travail | I have to move (out) for work | |
| Et, je n’ai <<que quelques jours de>> vacances per an. | And, I don’t have <<many>> vacations each year | |
| Alors on ne rompt pas? | Then we don’t break up | |
| discute | discuss | discuter (v) |
| Si <<vous vous retournez>>, je vais prendre une photo | See above | |
| La réunion <<vient de commencer>> | The meeting <<is about to start>> | |
| J’<<ai raison>>, et tu <<as tort>> | I <<am right>>, and tu <<are wrong>> | |
| Elle va grandir>> | She is going <<to grow up>> | grandir: grow, grow up |
| Je vais le remplir avec de l’eau | I am going <<to fill it>> (the glass) with water | |
| Il fait choisir <<la bonne taille>> | One must chose <<the right size>> | |
| Il est <<neuf>> | It is <<new>> | neuf (adj): nine, new, fresh, brand new, mint condition |
| Je voudrais des <<vêtements neufs>> | I would like new clothes | |
| Ce n’est pas <<la bonne taille>> | That’s not <<the right size>> | |
| J’ai choisi ces chaussures parce qu’elles sont <<bon marché>> | I chose these shoes because they are <<inexpensive>> | Compared to saying “pas cher” or “not expensive” |
| J’<<ai grandi>> à Californie | I <<grew up>> in California | |
In late March 2025, Google held a generative AI 5-day intensive for developers. It was hosted by Kaggle and included daily codelabs for the first four days. The course included daily live video sessions on YouTube with insights from current Google employees.
The course featured:
The course is targeted towards developers who already know how to write code. If you’re new to coding, the code will be more challenging because the codelabs require you to read the provided code.
It taught us how to use Vertex AI and Gemini APIs to implement generative AI for a range of use cases. We learned about agents, functions, creating custom models, how to fine tune a model, to name a few. The course is structured to provide coding examples that developers can use to develop there own generative AI solutions.
To take a future offering of the course, check this page for when another session will be offered. The first sesson was offered in November 2024 and the second was held in April 2025.