Meltemi 7B Instruct V1 by ilsp

 ยป  All LLMs  ยป  ilsp  ยป  Meltemi 7B Instruct V1   URL Share it on

  Arxiv:1803.05457   Arxiv:1905.07830   Arxiv:2009.03300   Arxiv:2109.07958   Arxiv:2308.16884   Arxiv:2407.20743   Autotrain compatible   Conversational   El   En   Endpoints compatible   Finetuned   Instruct   Mistral   Region:us   Safetensors   Sharded   Tensorflow

Meltemi 7B Instruct V1 Benchmarks

nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Meltemi 7B Instruct V1 (ilsp/Meltemi-7B-Instruct-v1)

Meltemi 7B Instruct V1 Parameters and Internals

Model Type 
text-generation
Use Cases 
Areas:
research, language understanding, Greek applications
Limitations:
might generate misleading, harmful, and toxic content.
Considerations:
Not aligned with human preferences
Additional Notes 
This model has a newer version available (v1.5) with enhancements.
Supported Languages 
el (fully supported), en (partially supported)
Training Details 
Data Sources:
Open-Platypus, Evol-Instruct, Capybara, Hand-crafted Greek dataset
Data Volume:
100k Greek machine translated instructions
Methodology:
Fine-tuning with Greek machine translated instructions; vocabulary extension
Context Length:
8192
Hardware Used:
Amazon Web Services via GRNET
Model Architecture:
Mistral-7b tokenizer with Greek tokens
Input Output 
Input Format:
<|system|> ... <|user|> ... <|assistant|> ...
Accepted Modalities:
text
Output Format:
Text generated in response
Performance Tips:
Ensure the BOS token is always included.
LLM NameMeltemi 7B Instruct V1
Repository ๐Ÿค—https://huggingface.co/ilsp/Meltemi-7B-Instruct-v1 
Model Size7b
Required VRAM15 GB
Updated2025-02-14
Maintainerilsp
Model Typemistral
Instruction-BasedYes
Model Files  10.0 GB: 1-of-2   5.0 GB: 2-of-2
Supported Languagesel en
Model ArchitectureMistralForCausalLM
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.37.2
Tokenizer ClassLlamaTokenizer
Padding Token<pad>
Vocabulary Size61520
Torch Data Typebfloat16

Quantized Models of the Meltemi 7B Instruct V1

Model
Likes
Downloads
VRAM
Meltemi 7B Instruct V1 AWQ61564 GB

Best Alternatives to Meltemi 7B Instruct V1

Best Alternatives
Context / RAM
Downloads
Likes
...Nemo Instruct 2407 Abliterated1000K / 24.5 GB517610
SpydazWeb AI HumanAI RP512K / 14.4 GB91
SpydazWeb AI HumanAI 002512K / 14.4 GB181
...daz Web AI ChatML 512K Project512K / 14.5 GB120
... Summarize 64K QLoRANET Merged128K / 4.1 GB860
...1 Summarize 64K LoRANET Merged128K / 14.4 GB150
Mistral 7B Instruct V0.232K / 14.4 GB33160952630
Mistral 7B Instruct V0.132K / 14.4 GB1560071562
...ity Instruct 7M Gen Mistral 7B32K / 14.4 GB36495
...ty Instruct 3M 0613 Mistral 7B32K / 14.4 GB617211
Note: green Score (e.g. "73.2") means that the model is better than ilsp/Meltemi-7B-Instruct-v1.

Rank the Meltemi 7B Instruct V1 Capabilities

๐Ÿ†˜ Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! ๐ŸŒŸ

Instruction Following and Task Automation  
Factuality and Completeness of Knowledge  
Censorship and Alignment  
Data Analysis and Insight Generation  
Text Generation  
Text Summarization and Feature Extraction  
Code Generation  
Multi-Language Support and Translation  

What open-source LLMs or SLMs are you in search of? 43106 in total.

Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241227