kcaverly/nous-capybara-34b-gguf – Run with an API on Replicate

kcaverly / nous-capybara-34b-gguf

A SOTA Nous Research finetune of 200k Yi-34B fine tuned on the Capybara dataset.

Cold

Public
1.9K runs
L40S
GitHub
Paper

Iterate in playground

Run with an API

Playground API Examples README Versions

Input

prompt

*string

Shift + Return to add a new line

Why does a GPU process matrix multiplication faster than a CPU?Why does a GPU process matrix multiplication faster than a CPU?

Instruction for model

system_prompt

string

Shift + Return to add a new line

You are 'Nous-Capybara', an AI assistant and your purpose and drive is to assist the user with any request they have.You are 'Nous-Capybara', an AI assistant and your purpose and drive is to assist the user with any request they have.

System prompt for the model, helps guides model behaviour.

Default: "You are 'Nous-Capybara', an AI assistant and your purpose and drive is to assist the user with any request they have."

prompt_template

string

Shift + Return to add a new line

SYSTEM: {system_prompt}
USER: {prompt}
ASSISTANT: SYSTEM: {system_prompt}
USER: {prompt}
ASSISTANT: 

Template to pass to model. Override if you are providing multi-turn instructions.

Default: "SYSTEM: {system_prompt}\nUSER: {prompt}\nASSISTANT: "

max_new_tokens

integer

Maximum new tokens to generate.

Default: -1

repeat_penalty

number

This parameter plays a role in controlling the behavior of an AI language model during conversation or text generation. Its purpose is to discourage the model from repeating itself too often by increasing the likelihood of following up with different content after each response. By adjusting this parameter, users can influence the model's tendency to either stay within familiar topics (lower penalty) or explore new ones (higher penalty). For instance, setting a high repeat penalty might result in more varied and dynamic conversations, whereas a low penalty could be suitable for scenarios where consistency and predictability are preferred.

Default: 1.1

temperature

number

This parameter used to control the 'warmth' or responsiveness of an AI model based on the LLaMA architecture. It adjusts how likely the model is to generate new, unexpected information versus sticking closely to what it has been trained on. A higher value for this parameter can lead to more creative and diverse responses, while a lower value results in safer, more conservative answers that are closer to those found in its training data. This parameter is particularly useful when fine-tuning models for specific tasks where you want to balance between generating novel insights and maintaining accuracy and coherence.

Default: 0.5

Run this model in Node.js with one line of code:

npx create-replicate --model=kcaverly/nous-capybara-34b-gguf

or set up a project from scratch

Install Replicate’s Node.js client library:

npm install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import and set up the client:

import Replicate from "replicate";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,
});

Run kcaverly/nous-capybara-34b-gguf using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "kcaverly/nous-capybara-34b-gguf:6b9b7741e719899f26571567b892a8900d0b517bfadee3997f5d477897d10eef",
  {
    input: {
      prompt: "Why does a GPU process matrix multiplication faster than a CPU?",
      temperature: 0.5,
      system_prompt: "You are 'Nous-Capybara', an AI assistant and your purpose and drive is to assist the user with any request they have.",
      max_new_tokens: -1,
      repeat_penalty: 1.1,
      prompt_template: "SYSTEM: {system_prompt}\nUSER: {prompt}\nASSISTANT: "
    }
  }
);

console.log(output);

To learn more, take a look at the guide on getting started with Node.js.

Install Replicate’s Python client library:

pip install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import the client:

import replicate

Run kcaverly/nous-capybara-34b-gguf using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "kcaverly/nous-capybara-34b-gguf:6b9b7741e719899f26571567b892a8900d0b517bfadee3997f5d477897d10eef",
    input={
        "prompt": "Why does a GPU process matrix multiplication faster than a CPU?",
        "temperature": 0.5,
        "system_prompt": "You are 'Nous-Capybara', an AI assistant and your purpose and drive is to assist the user with any request they have.",
        "max_new_tokens": -1,
        "repeat_penalty": 1.1,
        "prompt_template": "SYSTEM: {system_prompt}\nUSER: {prompt}\nASSISTANT: "
    }
)

# The kcaverly/nous-capybara-34b-gguf model can stream output as it's running.
# The predict method returns an iterator, and you can iterate over that output.
for item in output:
    # https://replicate.com/kcaverly/nous-capybara-34b-gguf/api#output-schema
    print(item, end="")

To learn more, take a look at the guide on getting started with Python.

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Run kcaverly/nous-capybara-34b-gguf using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

curl -s -X POST \
  -H "Authorization: Bearer $REPLICATE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Prefer: wait" \
  -d $'{
    "version": "kcaverly/nous-capybara-34b-gguf:6b9b7741e719899f26571567b892a8900d0b517bfadee3997f5d477897d10eef",
    "input": {
      "prompt": "Why does a GPU process matrix multiplication faster than a CPU?",
      "temperature": 0.5,
      "system_prompt": "You are \'Nous-Capybara\', an AI assistant and your purpose and drive is to assist the user with any request they have.",
      "max_new_tokens": -1,
      "repeat_penalty": 1.1,
      "prompt_template": "SYSTEM: {system_prompt}\\nUSER: {prompt}\\nASSISTANT: "
    }
  }' \
  https://api.replicate.com/v1/predictions

To learn more, take a look at Replicate’s HTTP API reference docs.

Output

1. Parallel Processing: GPUs are designed specifically for parallel processing, which makes them ideal for tasks like matrix multiplication that can be split into smaller, independent parts. On the other hand, CPUs are optimized for sequential processing and handling complex calculations. 2. Specialized Hardware: GPUs have specialized hardware called streaming multiprocessors that perform mathematical operations much faster than a CPU's cores. These streaming multiprocessors work together to process large amounts of data simultaneously, leading to significant performance improvements in tasks like matrix multiplication. 3. Memory Bandwidth: GPUs have higher memory bandwidth compared to CPUs, which allows them to load and manipulate larger datasets more efficiently. This is crucial for matrix multiplication as the operation requires a significant amount of memory to store intermediate results. 4. Compute Unified Device Architecture (CUDA) : NVIDIA's CUDA technology enables developers to write programs that can take advantage of GPUs' parallel processing capabilities. This allows CPU-intensive tasks like matrix multiplication to be offloaded to the GPU, resulting in faster computation times. 5. Task Switching: GPUs excel at performing the same task repeatedly and can switch between tasks much faster than a CPU. When performing matrix multiplication, this means that the GPU can work on multiple matrices simultaneously without wasting time on context switching.</s>

{
  "completed_at": "2024-01-09T19:51:45.485007Z",
  "created_at": "2024-01-09T19:47:31.371314Z",
  "data_removed": false,
  "error": null,
  "id": "imnbvg3bsfeqqtih6sfixxishu",
  "input": {
    "prompt": "Why does a GPU process matrix multiplication faster than a CPU?",
    "temperature": 0.5,
    "system_prompt": "You are 'Nous-Capybara', an AI assistant and your purpose and drive is to assist the user with any request they have.",
    "max_new_tokens": -1,
    "repeat_penalty": 1.1,
    "prompt_template": "SYSTEM: {system_prompt}\nUSER: {prompt}\nASSISTANT: "
  },
  "logs": "llama_print_timings:        load time =     773.84 ms\nllama_print_timings:      sample time =      83.54 ms /   266 runs   (    0.31 ms per token,  3184.18 tokens per second)\nllama_print_timings: prompt eval time =     770.88 ms /    56 tokens (   13.77 ms per token,    72.64 tokens per second)\nllama_print_timings:        eval time =   16066.31 ms /   265 runs   (   60.63 ms per token,    16.49 tokens per second)\nllama_print_timings:       total time =   18610.34 ms",
  "metrics": {
    "predict_time": 18.625253,
    "total_time": 254.113693
  },
  "output": [
    "1",
    ".",
    " Par",
    "allel",
    " Processing",
    ":",
    " GPUs",
    " are",
    " designed",
    " specifically",
    " for",
    " parallel",
    " processing",
    ",",
    " which",
    " makes",
    " them",
    " ideal",
    " for",
    " tasks",
    " like",
    " matrix",
    " multiplication",
    " that",
    " can",
    " be",
    " split",
    " into",
    " smaller",
    ",",
    " independent",
    " parts",
    ".",
    " On",
    " the",
    " other",
    " hand",
    ",",
    " CPU",
    "s",
    " are",
    " optimized",
    " for",
    " sequential",
    " processing",
    " and",
    " handling",
    " complex",
    " calculations",
    ".",
    "\n",
    "2",
    ".",
    " Special",
    "ized",
    " Hardware",
    ":",
    " GPUs",
    " have",
    " specialized",
    " hardware",
    " called",
    " streaming",
    " multi",
    "process",
    "ors",
    " that",
    " perform",
    " mathematical",
    " operations",
    " much",
    " faster",
    " than",
    " a",
    " CPU",
    "'",
    "s",
    " cores",
    ".",
    " These",
    " streaming",
    " multi",
    "process",
    "ors",
    " work",
    " together",
    " to",
    " process",
    " large",
    " amounts",
    " of",
    " data",
    " simultaneously",
    ",",
    " leading",
    " to",
    " significant",
    " performance",
    " improvements",
    " in",
    " tasks",
    " like",
    " matrix",
    " multiplication",
    ".",
    "\n",
    "3",
    ".",
    " Memory",
    " Band",
    "width",
    ":",
    " GPUs",
    " have",
    " higher",
    " memory",
    " bandwidth",
    " compared",
    " to",
    " CPU",
    "s",
    ",",
    " which",
    " allows",
    " them",
    " to",
    " load",
    " and",
    " manipulate",
    " larger",
    " datasets",
    " more",
    " efficiently",
    ".",
    " This",
    " is",
    " crucial",
    " for",
    " matrix",
    " multiplication",
    " as",
    " the",
    " operation",
    " requires",
    " a",
    " significant",
    " amount",
    " of",
    " memory",
    " to",
    " store",
    " intermediate",
    " results",
    ".",
    "\n",
    "4",
    ".",
    " Compute",
    " Unified",
    " Device",
    " Architecture",
    " (",
    "CU",
    "DA",
    ")",
    " :",
    " NVIDIA",
    "'",
    "s",
    " CU",
    "DA",
    " technology",
    " enables",
    " developers",
    " to",
    " write",
    " programs",
    " that",
    " can",
    " take",
    " advantage",
    " of",
    " GPUs",
    "'",
    " parallel",
    " processing",
    " capabilities",
    ".",
    " This",
    " allows",
    " CPU",
    "-",
    "int",
    "ensive",
    " tasks",
    " like",
    " matrix",
    " multiplication",
    " to",
    " be",
    " off",
    "loaded",
    " to",
    " the",
    " GPU",
    ",",
    " resulting",
    " in",
    " faster",
    " computation",
    " times",
    ".",
    "\n",
    "\n",
    "5",
    ".",
    " Task",
    " Sw",
    "itching",
    ":",
    " GPUs",
    " excel",
    " at",
    " performing",
    " the",
    " same",
    " task",
    " repeatedly",
    " and",
    " can",
    " switch",
    " between",
    " tasks",
    " much",
    " faster",
    " than",
    " a",
    " CPU",
    ".",
    " When",
    " performing",
    " matrix",
    " multiplication",
    ",",
    " this",
    " means",
    " that",
    " the",
    " GPU",
    " can",
    " work",
    " on",
    " multiple",
    " matrices",
    " simultaneously",
    " without",
    " wasting",
    " time",
    " on",
    " context",
    " switching",
    ".",
    "</",
    "s",
    ">",
    ""
  ],
  "started_at": "2024-01-09T19:51:26.859754Z",
  "status": "succeeded",
  "urls": {
    "stream": "https://streaming-api.svc.us.c.replicate.net/v1/predictions/imnbvg3bsfeqqtih6sfixxishu",
    "get": "https://api.replicate.com/v1/predictions/imnbvg3bsfeqqtih6sfixxishu",
    "cancel": "https://api.replicate.com/v1/predictions/imnbvg3bsfeqqtih6sfixxishu/cancel"
  },
  "version": "6b9b7741e719899f26571567b892a8900d0b517bfadee3997f5d477897d10eef"
}

Generated in

18.6 seconds

Tweak it Report

Run time and cost

This model costs approximately $2.22 to run on Replicate, or 0 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker.

This model runs on Nvidia L40S GPU hardware. Predictions typically complete within 38 minutes. The predict time for this model varies significantly based on the inputs.

Readme

Nous Capybara 34B is a SOTA Yi-34b 200k Fine tuned on the Capybara dataset. While the entire model can handle 200k, this version has been restricted to 32k for memory reasons.

The original model can be found here. The quantized version used is Q4_0, and can be found here.