---
license: cc-by-nc-4.0
base_model: google/gemma-2b
model-index:
- name: Octopus-V2-2B
  results: []
tags:
- function calling
- on-device language model
- android
inference: false
space: false
spaces: false
language:
- en
---

# Quantized Octopus V2: On-device language model for super agent

This repo includes two types of quantized models: **GGUF** and **AWQ**, for our Octopus V2 model at [NexaAIDev/Octopus-v2](https://huggingface.co./NexaAIDev/Octopus-v2)

<p align="center" width="100%">
  <a><img src="Octopus-logo.jpeg" alt="nexa-octopus" style="width: 40%; min-width: 300px; display: block; margin: auto;"></a>
</p>


# GGUF Qauntization
Run with [Ollama](https://github.com/ollama/ollama)

```bash
ollama run NexaAIDev/octopus-v2-Q4_K_M
```

# AWQ Quantization
Python example:

```python
from awq import AutoAWQForCausalLM
from transformers import AutoTokenizer, GemmaForCausalLM
import torch
import time
import numpy as np

def inference(input_text):

    tokens = tokenizer(
        input_text,
        return_tensors='pt'
    ).input_ids.cuda()

    start_time = time.time()
    generation_output = model.generate(
        tokens,
        do_sample=True,
        temperature=0.7,
        top_p=0.95,
        top_k=40,
        max_new_tokens=512
    )
    end_time = time.time()

    res = tokenizer.decode(generation_output[0])
    res = res.split(input_text)
    latency = end_time - start_time
    output_tokens = tokenizer.encode(res)
    num_output_tokens = len(output_tokens)
    throughput = num_output_tokens / latency

    return {"output": res[-1], "latency": latency, "throughput": throughput}


model_id = "path/to/Octopus-v2-AWQ"
model = AutoAWQForCausalLM.from_quantized(model_id, fuse_layers=True,
                                          trust_remote_code=False, safetensors=True)
tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=False)

prompts = ["Below is the query from the users, please call the correct function and generate the parameters to call the function.\n\nQuery: Can you take a photo using the back camera and save it to the default location? \n\nResponse:"]

avg_throughput = []
for prompt in prompts:
    out = inference(prompt)
    avg_throughput.append(out["throughput"])
    print("nexa model result:\n", out["output"])

print("avg throughput:", np.mean(avg_throughput))
```

# Quantized GGUF & AWQ Models Benchmark

| Name                   | Quant method | Bits | Size     | Response (t/s) | Use Cases                           |
| ---------------------- | ------------ | ---- | -------- | -------------- | ----------------------------------- |
| Octopus-v2-AWQ         | AWQ          | 4    | 3.00 GB  | 63.83          | fast, high quality, recommended     |
| Octopus-v2-Q2_K.gguf   | Q2_K         | 2    | 1.16 GB  | 57.81          | fast but high loss, not recommended |
| Octopus-v2-Q3_K.gguf   | Q3_K         | 3    | 1.38 GB  | 57.81          | extremely not recommended           |
| Octopus-v2-Q3_K_S.gguf | Q3_K_S       | 3    | 1.19 GB  | 52.13          | extremely not recommended           |
| Octopus-v2-Q3_K_M.gguf | Q3_K_M       | 3    | 1.38 GB  | 58.67          | moderate loss, not very recommended |
| Octopus-v2-Q3_K_L.gguf | Q3_K_L       | 3    | 1.47 GB  | 56.92          | not very recommended                |
| Octopus-v2-Q4_0.gguf   | Q4_0         | 4    | 1.55 GB  | 68.80          | moderate speed, recommended         |
| Octopus-v2-Q4_1.gguf   | Q4_1         | 4    | 1.68 GB  | 68.09          | moderate speed, recommended         |
| Octopus-v2-Q4_K.gguf   | Q4_K         | 4    | 1.63 GB  | 64.70          | moderate speed, recommended         |
| Octopus-v2-Q4_K_S.gguf | Q4_K_S       | 4    | 1.56 GB  | 62.16          | fast and accurate, very recommended |
| Octopus-v2-Q4_K_M.gguf | Q4_K_M       | 4    | 1.63 GB  | 64.74          | fast, recommended                   |
| Octopus-v2-Q5_0.gguf   | Q5_0         | 5    | 1.80 GB  | 64.80          | fast, recommended                   |
| Octopus-v2-Q5_1.gguf   | Q5_1         | 5    | 1.92 GB  | 63.42          | very big, prefer Q4                 |
| Octopus-v2-Q5_K.gguf   | Q5_K         | 5    | 1.84 GB  | 61.28          | big, recommended                    |
| Octopus-v2-Q5_K_S.gguf | Q5_K_S       | 5    | 1.80 GB  | 62.16          | big, recommended                    |
| Octopus-v2-Q5_K_M.gguf | Q5_K_M       | 5    | 1.71 GB  | 61.54          | big, recommended                    |
| Octopus-v2-Q6_K.gguf   | Q6_K         | 6    | 2.06 GB  | 55.94          | very big, not very recommended      |
| Octopus-v2-Q8_0.gguf   | Q8_0         | 8    | 2.67 GB  | 56.35          | very big, not very recommended      |
| Octopus-v2-f16.gguf    | f16          | 16   | 5.02 GB  | 36.27          | extremely big                       |
| Octopus-v2.gguf        |              |      | 10.00 GB |                |                                     |

_Quantized with llama.cpp_


**Acknowledgement**:  
We sincerely thank our community members, [Mingyuan](https://huggingface.co./ThunderBeee), [Zoey](https://huggingface.co./ZY6), [Brian](https://huggingface.co./JoyboyBrian), [Perry](https://huggingface.co./PerryCheng614), [Qi](https://huggingface.co./qiqiWav), [David](https://huggingface.co./Davidqian123) for their extraordinary contributions to this quantization effort.