File size: 2,988 Bytes
bb31303 b1d5ece b90e21e b1d5ece bb31303 b90e21e ad43e2e b90e21e ad43e2e b90e21e ad43e2e b90e21e ad43e2e b90e21e ad43e2e b90e21e bb31303 b90e21e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 |
---
base_model:
- mistralai/Ministral-8B-Instruct-2410
language:
- en
- fr
- de
- es
- it
- pt
- zh
- ja
- ru
- ko
license: other
license_name: mrl
license_link: https://mistral.ai/licenses/MRL-0.1.md
inference: false
---
## Installation
To use this model, install the required packages:
```bash
pip install -U mistral-common transformers torch
```
## Usage Example
Here's a Python script demonstrating how to use the model for chat completion:
```python
import torch
from pathlib import Path
from transformers import AutoModelForCausalLM
from mistral_common.tokens.tokenizers.mistral import MistralTokenizer
from mistral_common.tokens.tokenizers.tekken import SpecialTokenPolicy
from mistral_common.protocol.instruct.messages import UserMessage
from mistral_common.protocol.instruct.request import ChatCompletionRequest
from huggingface_hub import snapshot_download
def get_model_path(path_or_hf_repo: str, revision: str = None) -> Path:
"""Ensures the model is available locally, downloading if necessary."""
model_path = Path(path_or_hf_repo)
if not model_path.exists():
model_path = Path(
snapshot_download(
repo_id=path_or_hf_repo,
revision=revision,
allow_patterns=[
"*.json",
"*.safetensors",
"*.py",
"tokenizer.model",
"*.tiktoken",
"*.txt",
],
resume_download=True,
)
)
return model_path
def load_chat_request(message: str) -> ChatCompletionRequest:
"""Creates a chat completion request with a user message."""
return ChatCompletionRequest(messages=[UserMessage(content=message)])
# Model setup
model_name = "prince-canuma/Ministral-8B-Instruct-2410-HF"
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto")
model_path = get_model_path(model_name)
tokenizer = MistralTokenizer.from_file(f"{model_path}/tokenizer.json")
tekken = tokenizer.instruct_tokenizer.tokenizer
tekken.special_token_policy = SpecialTokenPolicy.IGNORE
# Chat interaction
user_message = "Tell me a short story about a robot learning to paint."
completion_request = load_chat_request(user_message)
tokens = tokenizer.encode_chat_completion(completion_request).tokens
input_ids = torch.tensor(tokens).unsqueeze(0)
# Generate response
output = model.generate(input_ids, max_new_tokens=500, temperature=0.7, do_sample=True)
response = tokenizer.decode(output[0][input_ids.shape[1]:].tolist())
print("User:", user_message)
print("Model:", response)
```
## Model Details
- **Developed by:** Mistral AI
- **Model type:** Causal Language Model
- **Language(s):** English
- **License:** [mrl](https://mistral.ai/licenses/MRL-0.1.md)
- **Resources for more information:**
- [Model Repository](https://huggingface.co/prince-canuma/Ministral-8B-Instruct-2410-HF)
- [Mistral AI GitHub](https://github.com/mistralai) |