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

Synthia 70B by migtissera

 »  All LLMs  »  migtissera  »  Synthia 70B   URL Share it on

  448   
  Arxiv:2302.13971   Arxiv:2306.02707   Autotrain compatible   En   Endpoints compatible   Fp16   Llama   Pytorch   Quantized   Region:us   Sharded
Model Card on HF 🤗: https://huggingface.co/migtissera/Synthia-70B 

Synthia 70B Benchmarks

ARC: 69.45 vs 96.7 (so35)-28.2%
HellaSwag: 87.11 vs 95.3 (gpt4)-8.6%
MMLU: 68.91 vs 88.3 (so35)-22%
TruthfulQA: 59.79 vs 59 (gpt4)1.3%
WinoGrande: 83.66 vs 87.5 (gpt4)-4.4%
GSM8K: 31.39 vs 96.4 (so35)-67.4%
LLME Score: 0.13943
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Synthia 70B (migtissera/Synthia-70B)

Synthia 70B Parameters and Internals

Model Type 
text-generation
Additional Notes 
This model is unregulated, which means it doesn't have restraints against generating inappropriate, biased, or offensive content. Exercise caution and cross-check information if necessary.
Training Details 
Data Sources:
Orca style datasets
Methodology:
fine-tuned for instruction following and long-form conversations
Input Output 
Input Format:
SYSTEM: {system message} USER: {user message} ASSISTANT: {response}
Accepted Modalities:
text
Output Format:
text
LLM NameSynthia 70B
Repository 🤗https://huggingface.co/migtissera/Synthia-70B 
Model Size70b
Required VRAM138 GB
Updated2025-04-24
Maintainermigtissera
Model Typellama
Model Files  9.8 GB: 1-of-15   9.8 GB: 2-of-15   10.0 GB: 3-of-15   9.8 GB: 4-of-15   9.8 GB: 5-of-15   9.8 GB: 6-of-15   10.0 GB: 7-of-15   9.8 GB: 8-of-15   9.8 GB: 9-of-15   9.8 GB: 10-of-15   10.0 GB: 11-of-15   9.8 GB: 12-of-15   9.8 GB: 13-of-15   9.5 GB: 14-of-15   0.5 GB: 15-of-15
Supported Languagesen
Quantization Typefp16
Model ArchitectureLlamaForCausalLM
Licensellama2
Context Length2048
Model Max Length2048
Transformers Version4.32.0.dev0
Tokenizer ClassLlamaTokenizer
Beginning of Sentence Token<s>
End of Sentence Token</s>
Unk Token<unk>
Vocabulary Size32000
Torch Data Typefloat16

Quantized Models of the Synthia 70B

Model
Likes
Downloads
VRAM
Synthia 70B GGUF782329 GB
Synthia 70B AWQ1536 GB
Synthia 70B GPTQ2535 GB
Synthia 70B GGML2528 GB

Best Alternatives to Synthia 70B

Best Alternatives
Context / RAM
Downloads
Likes
...B Instruct Gradient 1048K 8bit1024K / 75 GB25
...B Instruct Gradient 1048K 4bit1024K / 39.7 GB33
...B Instruct Gradient 1048K 2bit1024K / 21.9 GB22
...0B Instruct Gradient 262K 8bit256K / 75 GB102
...0B Instruct Gradient 262K 4bit256K / 39.7 GB23
...0B Instruct Gradient 262K 2bit256K / 21.9 GB21
... Gradient 262K 2.25bpw H6 EXL2256K / 22.2 GB60
...t Gradient 262K 4.0bpw H6 EXL2256K / 37.2 GB21
...t Gradient 262K 2.4bpw H6 EXL2256K / 23.5 GB50
...t Gradient 262K 3.5bpw H6 EXL2256K / 32.9 GB50
Note: green Score (e.g. "73.2") means that the model is better than migtissera/Synthia-70B.

Rank the Synthia 70B 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? 46635 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 v20241227
  Support LLM Explorer