LLM EXPLORER 63,407 MODELS INDEXED

Tulu 2 DPO 70B 2.4bpw H6 EXL2 by LoneStriker

By LoneStriker · 5 downloads
HF Repo ↗ Copy URL

Tulu 2 DPO 70B 2.4bpw H6 EXL2 is an open-source language model by LoneStriker. Features: 70b LLM, VRAM: 21.3GB, Context: 8K, Quantized, LLM Explorer Score: 0.09.

  Arxiv:2305.18290   Arxiv:2311.10702 Base model:finetune:meta-llama... Base model:meta-llama/llama-2-... Dataset:allenai/tulu-v2-sft-mi... Dataset:huggingfaceh4/ultrafee...   En   Endpoints compatible   Exl2   Llama   Pytorch   Quantized   Region:us   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/LoneStriker/tulu-2-dpo-70b-2.4bpw-h6-exl2 

Tulu 2 DPO 70B 2.4bpw H6 EXL2 Parameters and Internals

Model Type 
chat model, instruction role-tuned, RLHF
Use Cases 
Areas:
language modeling, NLP
Primary Use Cases:
instruction following, dialog assistance
Limitations:
may produce problematic outputs, not aligned to generate safe completions
Additional Notes 
Model not deployed with in-the-loop filtering.
Supported Languages 
en (Primarily)
Training Details 
Data Sources:
HuggingFaceH4/ultrafeedback_binarized, allenai/tulu-v2-sft-mixture
Methodology:
Direct Preference Optimization (DPO)
Input Output 
Input Format:
<|user|> Your message here! <|assistant|> Output follows as AI response.
Accepted Modalities:
text
Output Format:
Generated response with alignment and improvements.
Performance Tips:
Include a newline after <|assistant|> for improved generation quality.
LLM NameTulu 2 DPO 70B 2.4bpw H6 EXL2
Repository 🤗https://huggingface.co/LoneStriker/tulu-2-dpo-70b-2.4bpw-h6-exl2 
Base Model(s)  Llama 2 70B Hf   meta-llama/Llama-2-70b-hf
Model Size70b
Required VRAM21.3 GB
Updated2026-08-07
MaintainerLoneStriker
Model Typellama
Model Files  8.6 GB: 1-of-3   8.6 GB: 2-of-3   4.1 GB: 3-of-3
Supported Languagesen
Quantization Typeexl2
Model ArchitectureLlamaForCausalLM
Context Length8192
Model Max Length8192
Transformers Version4.33.2
Tokenizer ClassLlamaTokenizer
Beginning of Sentence Token<s>
End of Sentence Token</s>
Unk Token<unk>
Vocabulary Size32000
Torch Data Typebfloat16

Best Alternatives to Tulu 2 DPO 70B 2.4bpw H6 EXL2

Best Alternatives
Context / RAM
Downloads
Likes
...B Instruct Gradient 1048K 8bit1024K / 75 GB415
...B Instruct Gradient 1048K 4bit1024K / 39.7 GB253
...B Instruct Gradient 1048K 2bit1024K / 21.9 GB252
...0B Instruct Gradient 262K 4bit256K / 39.7 GB253
...0B Instruct Gradient 262K 8bit256K / 75 GB262
...0B Instruct Gradient 262K 2bit256K / 21.9 GB201
... Gradient 262K 2.25bpw H6 EXL2256K / 22.2 GB80
...t Gradient 262K 4.0bpw H6 EXL2256K / 37.2 GB71
...t Gradient 262K 2.4bpw H6 EXL2256K / 23.5 GB60
...t Gradient 262K 3.5bpw H6 EXL2256K / 32.9 GB60
Note: green Score (e.g. "73.2") means that the model is better than LoneStriker/tulu-2-dpo-70b-2.4bpw-h6-exl2.