umamarahman/tinyStories:f32

umamarahman/ tinyStories:f32

27 Downloads Updated 1 year ago

This is the 110M parameter Llama 2 architecture model trained on the TinyStories dataset

ollama run umamarahman/tinyStories:f32

curl http://localhost:11434/api/chat \
  -d '{
    "model": "umamarahman/tinyStories:f32",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='umamarahman/tinyStories:f32',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'umamarahman/tinyStories:f32',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 1 year ago

1 year ago

8141f1f8d048 · 537MB

model

archllama

parameters134M

quantizationF32

537MB

params

{ "num_ctx": 1024, "stop": [ "<|im_start|>", "<|im_end|>" ] }

74B

template

{{ if .System }}<|im_start|>system {{ .System }}<|im_end|> {{ end }}{{ if .Prompt }}<|im_start|>user

182B

Readme

This is the 110M parameter Llama 2 architecture model trained on the TinyStories dataset. These are converted from karpathy/tinyllamas. See the llama2.c project for more details.