Edit model card

nanoT5-mid-2k-instruct

This model is pszemraj/nanoT5-mid-65kBPE-2048 fine-tuned on ~2M examples to complete "instruct"-style tasks. It performs quite coherently for its size.

Details:

  • Pre-trained and fine-tuned with 2048 in/out context length
  • Variety of tasks in training data

Usage example

from transformers import pipeline

pipe = pipeline(
    "text2text-generation",
    model="pszemraj/nanoT5-mid-2k-instruct",
    device_map="auto",
)

prompt = "Write a Python code to reverse a linked list."
res = pipe(
    prompt,
    max_new_tokens=384,
    top_k=4,
    penalty_alpha=0.6,
    # no_repeat_ngram_size=6, # use for repeats
)
print(res[0]["generated_text"])
Click to see example response
Here's a Python code that reverses a linked list:

\```python
class Node:
    def __init__(self, data):
        self.data = data
        self.next = None

class LinkedList:
    def __init__(self):
        self.head = None

    def append(self, data):
        new_node = Node(data)
        new_node.next = Node(data)
        new_node.next = new_node

    def reverse(self):
        new_node = Node(data)
        new_node.next = self.head
        new_node.next = new_node

    def print_list(self):
        print(self.head)

# Example usage:
# Create a linked list
linked_list = LinkedList()
# Reverse the linked list
linked_list.reverse()
\```

In this code, we define a class `LinkedList` that represents a node in the linked list. The `__init__` method is used to initialize the head of the linked list.

The `reverse` method takes the head of the linked list as an argument and returns the reverse of the linked list.

The `reverse` method takes the head of the linked list as an argument and returns the reverse of the linked list.

The `reverse` method takes the head of the linked list as an argument and returns the reverse of the linked list.

In the example usage, we create a linked list `linked_list` with some data. We call the `reverse` method on the head of the linked list and print the result.

(the \ character was manually added before the code fences here for display reasons)

Downloads last month
50
Safetensors
Model size
637M params
Tensor type
F32
·
Inference API
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for pszemraj/nanoT5-mid-2k-instruct

Finetuned
this model

Space using pszemraj/nanoT5-mid-2k-instruct 1