Uploaded model

• Developed by: AdrienB134
• License: apache-2.0
• Finetuned from model : unsloth/mistral-7b-v0.3

This mistral model was trained 2x faster with Unsloth and Huggingface's TRL library.

How to use

from unsloth import FastLanguageModel
import torch

max_seq_length = 32_768 # Choose any! We auto support RoPE Scaling internally!
dtype = None # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+
load_in_4bit = False # Use 4bit quantization to reduce memory usage. Can be True.


model, tokenizer = FastLanguageModel.from_pretrained(
    model_name = "AdrienB134/French-Alpaca-Mistral-7B-v0.3", 
    max_seq_length = max_seq_length,
    dtype = dtype,
    load_in_4bit = load_in_4bit,
    # token = "hf_...", # use one if using gated models like meta-llama/Llama-2-7b-hf
)

alpaca_prompt = """Ci-dessous tu trouveras une instruction qui décrit une tâche, accompagnée d'un contexte qui donne plus d'informations. Ecrit une réponse appropriée à l'instruction.
### Instruction:
{}

### Contexte:
{}

### Response:
{}"""

FastLanguageModel.for_inference(model) # Enable native 2x faster inference
inputs = tokenizer(
[
    alpaca_prompt.format(
        "Continue la série de fibonacci.", # instruction
        "1, 1, 2, 3, 5, 8", # contexte
        "", # output - leave this blank for generation!
    )
], return_tensors = "pt").to("cuda")

from transformers import TextStreamer
text_streamer = TextStreamer(tokenizer)
_ = model.generate(**inputs, streamer = text_streamer, max_new_tokens = 128)