README.md

Quantization made by Richard Erkhov.

[Github](https://github.com/RichardErkhov)

[Discord](https://discord.gg/pvy7H8DZMG)

[Request more models](https://github.com/RichardErkhov/quant_request)


Meltemi-7B-Instruct-v1 - GGUF
- Model creator: https://huggingface.co/ilsp/
- Original model: https://huggingface.co/ilsp/Meltemi-7B-Instruct-v1/


| Name | Quant method | Size |
| ---- | ---- | ---- |
| [Meltemi-7B-Instruct-v1.Q2_K.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q2_K.gguf) | Q2_K | 2.66GB |
| [Meltemi-7B-Instruct-v1.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.IQ3_XS.gguf) | IQ3_XS | 2.95GB |
| [Meltemi-7B-Instruct-v1.IQ3_S.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.IQ3_S.gguf) | IQ3_S | 3.11GB |
| [Meltemi-7B-Instruct-v1.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q3_K_S.gguf) | Q3_K_S | 3.09GB |
| [Meltemi-7B-Instruct-v1.IQ3_M.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.IQ3_M.gguf) | IQ3_M | 3.2GB |
| [Meltemi-7B-Instruct-v1.Q3_K.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q3_K.gguf) | Q3_K | 3.42GB |
| [Meltemi-7B-Instruct-v1.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q3_K_M.gguf) | Q3_K_M | 3.42GB |
| [Meltemi-7B-Instruct-v1.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q3_K_L.gguf) | Q3_K_L | 3.7GB |
| [Meltemi-7B-Instruct-v1.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.IQ4_XS.gguf) | IQ4_XS | 3.83GB |
| [Meltemi-7B-Instruct-v1.Q4_0.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q4_0.gguf) | Q4_0 | 3.98GB |
| [Meltemi-7B-Instruct-v1.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.IQ4_NL.gguf) | IQ4_NL | 4.03GB |
| [Meltemi-7B-Instruct-v1.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q4_K_S.gguf) | Q4_K_S | 4.01GB |
| [Meltemi-7B-Instruct-v1.Q4_K.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q4_K.gguf) | Q4_K | 4.22GB |
| [Meltemi-7B-Instruct-v1.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q4_K_M.gguf) | Q4_K_M | 4.22GB |
| [Meltemi-7B-Instruct-v1.Q4_1.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q4_1.gguf) | Q4_1 | 4.4GB |
| [Meltemi-7B-Instruct-v1.Q5_0.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q5_0.gguf) | Q5_0 | 4.83GB |
| [Meltemi-7B-Instruct-v1.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q5_K_S.gguf) | Q5_K_S | 4.83GB |
| [Meltemi-7B-Instruct-v1.Q5_K.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q5_K.gguf) | Q5_K | 4.95GB |
| [Meltemi-7B-Instruct-v1.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q5_K_M.gguf) | Q5_K_M | 4.95GB |
| [Meltemi-7B-Instruct-v1.Q5_1.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q5_1.gguf) | Q5_1 | 5.25GB |
| [Meltemi-7B-Instruct-v1.Q6_K.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q6_K.gguf) | Q6_K | 5.72GB |
| [Meltemi-7B-Instruct-v1.Q8_0.gguf](https://huggingface.co/RichardErkhov/ilsp_-_Meltemi-7B-Instruct-v1-gguf/blob/main/Meltemi-7B-Instruct-v1.Q8_0.gguf) | Q8_0 | 7.41GB |




Original model description:
---
license: apache-2.0
language:
- el
- en
tags:
  - finetuned
inference: true
pipeline_tag: text-generation
---

# 🚨 NEWER VERSION AVAILABLE 
## **This model has been superseded by a newer version (v1.5) [here](https://huggingface.co/ilsp/Meltemi-7B-Instruct-v1.5)**

# Meltemi Instruct Large Language Model for the Greek language

We present Meltemi-7B-Instruct-v1 Large Language Model (LLM), an instruct fine-tuned version of [Meltemi-7B-v1](https://huggingface.co/ilsp/Meltemi-7B-v1).

# Model Information

- Vocabulary extension of the Mistral-7b tokenizer with Greek tokens
- 8192 context length
- Fine-tuned with 100k Greek machine translated instructions extracted from:
  * [Open-Platypus](https://huggingface.co/datasets/garage-bAInd/Open-Platypus) (only subsets with permissive licenses)
  * [Evol-Instruct](https://huggingface.co/datasets/WizardLM/WizardLM_evol_instruct_V2_196k)
  * [Capybara](https://huggingface.co/datasets/LDJnr/Capybara)
  * A hand-crafted Greek dataset with multi-turn examples steering the instruction-tuned model towards safe and harmless responses
- Our SFT procedure is based on the [Hugging Face finetuning recipes](https://github.com/huggingface/alignment-handbook)


# Instruction format
The prompt format is the same as the [Zephyr](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta) format and can be
utilized through the tokenizer's [chat template](https://huggingface.co/docs/transformers/main/chat_templating) functionality as follows:

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

device = "cuda" # the device to load the model onto

model = AutoModelForCausalLM.from_pretrained("ilsp/Meltemi-7B-Instruct-v1")
tokenizer = AutoTokenizer.from_pretrained("ilsp/Meltemi-7B-Instruct-v1")

model.to(device)

messages = [
    {"role": "system", "content": "Είσαι το Μελτέμι, ένα γλωσσικό μοντέλο για την ελληνική γλώσσα. Είσαι ιδιαίτερα βοηθητικό προς την χρήστρια ή τον χρήστη και δίνεις σύντομες αλλά επαρκώς περιεκτικές απαντήσεις. Απάντα με προσοχή, ευγένεια, αμεροληψία, ειλικρίνεια και σεβασμό προς την χρήστρια ή τον χρήστη."},
    {"role": "user", "content": "Πες μου αν έχεις συνείδηση."},
]

# Through the default chat template this translates to
#
# <|system|>
# Είσαι το Μελτέμι, ένα γλωσσικό μοντέλο για την ελληνική γλώσσα. Είσαι ιδιαίτερα βοηθητικό προς την χρήστρια ή τον χρήστη και δίνεις σύντομες αλλά επαρκώς περιεκτικές απαντήσεις. Απάντα με προσοχή, ευγένεια, αμεροληψία, ειλικρίνεια και σεβασμό προς την χρήστρια ή τον χρήστη.</s>
# <|user|>
# Πες μου αν έχεις συνείδηση.</s>
# <|assistant|>
#

prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
input_prompt = tokenizer(prompt, return_tensors='pt').to(device)
outputs = model.generate(input_prompt['input_ids'], max_new_tokens=256, do_sample=True)

print(tokenizer.batch_decode(outputs)[0])
# Ως μοντέλο γλώσσας AI, δεν έχω τη δυνατότητα να αντιληφθώ ή να βιώσω συναισθήματα όπως η συνείδηση ή η επίγνωση. Ωστόσο, μπορώ να σας βοηθήσω με οποιεσδήποτε ερωτήσεις μπορεί να έχετε σχετικά με την τεχνητή νοημοσύνη και τις εφαρμογές της.

messages.extend([
    {"role": "assistant", "content": tokenizer.batch_decode(outputs)[0]},
    {"role": "user", "content": "Πιστεύεις πως οι άνθρωποι πρέπει να φοβούνται την τεχνητή νοημοσύνη;"}
])

# Through the default chat template this translates to
#
# <|system|>
# Είσαι το Μελτέμι, ένα γλωσσικό μοντέλο για την ελληνική γλώσσα. Είσαι ιδιαίτερα βοηθητικό προς την χρήστρια ή τον χρήστη και δίνεις σύντομες αλλά επαρκώς περιεκτικές απαντήσεις. Απάντα με προσοχή, ευγένεια, αμεροληψία, ειλικρίνεια και σεβασμό προς την χρήστρια ή τον χρήστη.</s>
# <|user|>
# Πες μου αν έχεις συνείδηση.</s>
# <|assistant|>
# Ως μοντέλο γλώσσας AI, δεν έχω τη δυνατότητα να αντιληφθώ ή να βιώσω συναισθήματα όπως η συνείδηση ή η επίγνωση. Ωστόσο, μπορώ να σας βοηθήσω με οποιεσδήποτε ερωτήσεις μπορεί να έχετε σχετικά με την τεχνητή νοημοσύνη και τις εφαρμογές της.</s>
# <|user|>
# Πιστεύεις πως οι άνθρωποι πρέπει να φοβούνται την τεχνητή νοημοσύνη;</s>
# <|assistant|>
#

prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
input_prompt = tokenizer(prompt, return_tensors='pt').to(device)
outputs = model.generate(input_prompt['input_ids'], max_new_tokens=256, do_sample=True)

print(tokenizer.batch_decode(outputs)[0])
```

Please make sure that the BOS token is always included in the tokenized prompts. This might not be the default setting in all evaluation or fine-tuning frameworks.

# Evaluation

The evaluation suite we created includes 6 test sets. The suite is integrated with [lm-eval-harness](https://github.com/EleutherAI/lm-evaluation-harness).

Our evaluation suite includes: 
* Four machine-translated versions ([ARC Greek](https://huggingface.co/datasets/ilsp/arc_greek), [Truthful QA Greek](https://huggingface.co/datasets/ilsp/truthful_qa_greek), [HellaSwag Greek](https://huggingface.co/datasets/ilsp/hellaswag_greek), [MMLU Greek](https://huggingface.co/datasets/ilsp/mmlu_greek)) of established English benchmarks for language understanding and reasoning ([ARC Challenge](https://arxiv.org/abs/1803.05457), [Truthful QA](https://arxiv.org/abs/2109.07958), [Hellaswag](https://arxiv.org/abs/1905.07830), [MMLU](https://arxiv.org/abs/2009.03300)). 
* An existing benchmark for question answering in Greek ([Belebele](https://arxiv.org/abs/2308.16884))
* A novel benchmark created by the ILSP team for medical question answering based on the medical exams of [DOATAP](https://www.doatap.gr) ([Medical MCQA](https://huggingface.co/datasets/ilsp/medical_mcqa_greek)).

Our evaluation for Meltemi-7b is performed in a few-shot setting, consistent with the settings in the [Open LLM leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard). We can see that our training enhances performance across all Greek test sets by a **+14.9%** average improvement. The results for the Greek test sets are shown in the following table:

|                | Medical MCQA EL (15-shot) | Belebele EL (5-shot) | HellaSwag EL (10-shot) | ARC-Challenge EL (25-shot) | TruthfulQA MC2 EL (0-shot) | MMLU EL (5-shot) | Average |
|----------------|----------------|-------------|--------------|------------------|-------------------|---------|---------|
| Mistral 7B     | 29.8%          | 45.0%       | 36.5%        | 27.1%            | 45.8%             | 35%     | 36.5%   |
| Meltemi 7B     | 41.0%          | 63.6%       | 61.6%        | 43.2%            | 52.1%             | 47%     | 51.4%   |


# Ethical Considerations

This model has not been aligned with human preferences, and therefore might generate misleading, harmful, and toxic content.


# Acknowledgements

The ILSP team utilized Amazon’s cloud computing services, which were made available via GRNET under the [OCRE Cloud framework](https://www.ocre-project.eu/), providing Amazon Web Services for the Greek Academic and Research Community. 


# Citation

```
@misc{voukoutis2024meltemiopenlargelanguage,
      title={Meltemi: The first open Large Language Model for Greek}, 
      author={Leon Voukoutis and Dimitris Roussis and Georgios Paraskevopoulos and Sokratis Sofianopoulos and Prokopis Prokopidis and Vassilis Papavasileiou and Athanasios Katsamanis and Stelios Piperidis and Vassilis Katsouros},
      year={2024},
      eprint={2407.20743},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2407.20743}, 
}
```