keras_hub_quickstart.ipynb

Kernel: Python 3

KerasHub quickstart

We recommend running this example in Colab's GPU runtime. It will run on Jax, TensorFlow or PyTorch, simply change the line below.

import os
os.environ["KERAS_BACKEND"] = "jax"  # Or "torch" or "tensorflow"

import keras
import keras_hub

Ask questions with an instruction tuned Gemma checkpoint

causal_lm = keras_hub.models.CausalLM.from_preset(
    "gemma2_instruct_2b_en",
    dtype="float16",
)
causal_lm.summary()

prompt = """<start_of_turn>user
Write a short python program to print the first 100 primes.
<end_of_turn>
<start_of_turn>model
"""
text_output = causal_lm.generate(prompt, max_length=512)

print(text_output)

<start_of_turn>user
Write a short python program to print the first 100 primes.
<end_of_turn>
<start_of_turn>model
```python
def is_prime(num):
  """Checks if a number is prime."""
  if num <= 1:
    return False
  for i in range(2, int(num**0.5) + 1):
    if num % i == 0:
      return False
  return True

count = 0
num = 2
primes = []

while count < 100:
  if is_prime(num):
    primes.append(num)
    count += 1
  num += 1

print("The first 100 prime numbers are:", primes)
```

**Explanation:**

1. **`is_prime(num)` function:**
   - Takes an integer `num` as input.
   - Returns `False` if `num` is less than or equal to 1 (not prime).
   - Iterates from 2 to the square root of `num`. If `num` is divisible by any number in this range, it's not prime, so it returns `False`.
   - If the loop completes without finding a divisor, `num` is prime, and it returns `True`.

2. **Main loop:**
   - Initializes `count` to 0 (to keep track of how many primes we've found).
   - Initializes `num` to 2 (the first prime number).
   - Creates an empty list `primes` to store the prime numbers.
   - Enters a `while` loop that continues until `count` reaches 100 (we've found 100 primes).
   - Inside the loop:
     - Calls `is_prime(num)` to check if the current `num` is prime.
     - If it's prime:
       - Appends `num` to the `primes` list.
       - Increments `count`.
     - Increments `num` to check the next number.

3. **Output:**
   - After the loop, prints the `primes` list, which contains the first 100 prime numbers.


Let me know if you'd like to explore other prime number algorithms or have any more questions! 
<end_of_turn>

del causal_lm  # Free memory.

Generate images with Stable Diffusion 3

text_to_image = keras_hub.models.TextToImage.from_preset(
    "stable_diffusion_3_medium",
    dtype="float16",
)

Downloading from https://www.kaggle.com/api/v1/models/keras/stablediffusion3/keras/stable_diffusion_3_medium/3/download/config.json...

100%|██████████| 3.07k/3.07k [00:00<00:00, 2.45MB/s]

Downloading from https://www.kaggle.com/api/v1/models/keras/stablediffusion3/keras/stable_diffusion_3_medium/3/download/model.weights.h5...

100%|██████████| 5.57G/5.57G [04:57<00:00, 20.1MB/s]

Downloading from https://www.kaggle.com/api/v1/models/keras/stablediffusion3/keras/stable_diffusion_3_medium/3/download/preprocessor.json...

100%|██████████| 4.08k/4.08k [00:00<00:00, 3.50MB/s]

Downloading from https://www.kaggle.com/api/v1/models/keras/stablediffusion3/keras/stable_diffusion_3_medium/3/download/assets/clip_l_tokenizer/vocabulary.json...

100%|██████████| 976k/976k [00:01<00:00, 863kB/s]

Downloading from https://www.kaggle.com/api/v1/models/keras/stablediffusion3/keras/stable_diffusion_3_medium/3/download/assets/clip_l_tokenizer/merges.txt...

100%|██████████| 512k/512k [00:00<00:00, 571kB/s]

Downloading from https://www.kaggle.com/api/v1/models/keras/stablediffusion3/keras/stable_diffusion_3_medium/3/download/assets/clip_g_tokenizer/vocabulary.json...

100%|██████████| 976k/976k [00:01<00:00, 872kB/s]

Downloading from https://www.kaggle.com/api/v1/models/keras/stablediffusion3/keras/stable_diffusion_3_medium/3/download/assets/clip_g_tokenizer/merges.txt...

100%|██████████| 512k/512k [00:00<00:00, 576kB/s]

prompt = "Astronaut in a jungle, detailed"
image_output = text_to_image.generate(prompt)

from PIL import Image

Image.fromarray(image_output)