LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Tamil Mistral 7B Instruct V0.1 by Hemanth-thunder

 »  All LLMs  »  Hemanth-thunder  »  Tamil Mistral 7B Instruct V0.1   URL Share it on

  551   
  Autotrain compatible Base model:hemanth-thunder/tam... Base model:quantized:hemanth-t...   Chatml Dataset:hemanth-thunder/tamil-...   Distillation   Dpo   Endpoints compatible   Finetuned   Function calling   Gguf   Gpt4   Instruct   Json mode   Mistral   Quantized   Region:us   Rlhf   Safetensors   Sharded   Synthetic data   Ta   Tensorflow

Tamil Mistral 7B Instruct V0.1 Benchmarks

ARC: 27.39 vs 96.7 (so35)-71.7%
HellaSwag: 27.16 vs 95.3 (gpt4)-71.5%
MMLU: 24.42 vs 88.3 (so35)-72.3%
TruthfulQA: 47.27 vs 59 (gpt4)-19.9%
WinoGrande: 48.7 vs 87.5 (gpt4)-44.3%
LLME Score: 0.25296
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

Tamil Mistral 7B Instruct V0.1 Parameters and Internals

Model Type 
text generation, language understanding
Use Cases 
Areas:
research, text comprehension, instruction-following
Primary Use Cases:
instruction-based interactions in Tamil language
Limitations:
does not have any moderation mechanisms
Considerations:
Should be used in safe environments with necessary moderation.
Additional Notes 
The model can interact in the Tamil language, representing a significant advancement for non-English LLMs.
Supported Languages 
ta (advanced)
Training Details 
Data Sources:
Hemanth-thunder/tamil-open-instruct-v1
Data Volume:
400k instructions
Methodology:
instruction fine-tuning
Training Time:
18 hrs
Hardware Used:
NVIDIA RTX A6000 48GB with batch size of 30
Model Architecture:
Mistral-7B-v0.1 with Grouped-Query Attention, Sliding-Window Attention, Byte-fallback BPE tokenizer
Input Output 
Input Format:
Instruction, Input, and Response structured template
Accepted Modalities:
text
Output Format:
text
LLM NameTamil Mistral 7B Instruct V0.1
Repository 🤗https://huggingface.co/Hemanth-thunder/Tamil-Mistral-7B-Instruct-v0.1 
Base Model(s)  Hemanth-thunder/Tamil-Mistral-7B-v0.1   Hemanth-thunder/Tamil-Mistral-7B-v0.1
Model Size7b
Required VRAM14.8 GB
Updated2024-11-21
MaintainerHemanth-thunder
Model Typemistral
Instruction-BasedYes
Model Files  5.0 GB: 1-of-3   4.9 GB: 2-of-3   4.9 GB: 3-of-3   4.5 GB   5.2 GB
Supported Languagesta
GGUF QuantizationYes
Quantization Typegguf
Model ArchitectureMistralForCausalLM
Licenseapache-2.0
Context Length32768
Model Max Length32768
Transformers Version4.38.2
Tokenizer ClassLlamaTokenizer
Padding Token</s>
Vocabulary Size50622
Torch Data Typefloat16
Tamil Mistral 7B Instruct V0.1 (Hemanth-thunder/Tamil-Mistral-7B-Instruct-v0.1)

Best Alternatives to Tamil Mistral 7B Instruct V0.1

Best Alternatives
Context / RAM
Downloads
Likes
Mistral 7B Instruct V0.2.gguf32K / 14.5 GB103
Mistral 7B Instruct V0.2 GGUF32K / 4.4 GB532
...andle Mistral 7B Instruct V0.232K / 14.4 GB410
BioMistralMerged32K / 14.4 GB39030
...stral 7B V0.3 Chinese Chat F1632K / 14.5 GB6341
...tral 7B V0.3 Chinese Chat 4bit32K / 4.1 GB2271
Mistral 7B Instruct V0.3 GGUF32K / 2.7 GB930
Mistral 7B Instruct V0.3 GGUF32K / 2.7 GB490
...tral 7B V0.3 Chinese Chat 8bit32K / 7.7 GB163
...7B Instruct V0.2 SFT Local 27632K / 14.4 GB80
Note: green Score (e.g. "73.2") means that the model is better than Hemanth-thunder/Tamil-Mistral-7B-Instruct-v0.1.

Rank the Tamil Mistral 7B Instruct V0.1 Capabilities

🆘 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! 🌟

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? 38149 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241110
  Support LLM Explorer