LLM EXPLORER 61,491 MODELS INDEXED

Bagel DPO 34B V0.2 AWQ by TheBloke

By TheBloke · 10 downloads
HF Repo ↗ Copy URL

Bagel DPO 34B V0.2 AWQ is an open-source language model by TheBloke. Features: 34b LLM, VRAM: 19.3GB, Context: 195K, License: other, Quantized, LLM Explorer Score: 0.1.

  4-bit   Awq Base model:jondurbin/bagel-dpo... Base model:quantized:jondurbin...   Conversational   Dataset:ai2 arc Dataset:allenai/ultrafeedback ...   Dataset:boolq   Dataset:cais/mmlu   Dataset:cakiki/rosetta-code   Dataset:codeparrot/apps   Dataset:datasets/winogrande   Dataset:drop   Dataset:facebook/belebele   Dataset:intel/orca dpo pairs Dataset:jondurbin/cinematika-v... Dataset:jondurbin/truthy-dpo-v...   Dataset:julielab/emobank   Dataset:kingbri/pippa-sharegpt   Dataset:ldjnr/capybara   Dataset:lmsys/lmsys-chat-1m Dataset:migtissera/synthia-v1.... Dataset:muennighoff/natural-in...   Dataset:nvidia/helpsteer   Dataset:open-orca/slimorca   Dataset:openbookqa   Dataset:piqa   Dataset:spider   Dataset:squad v2 Dataset:squish42/bluemoon-fand...   Dataset:tiger-lab/mathinstruct   Dataset:unalignment/spicy-3.1 Dataset:unalignment/toxic-dpo-... Dataset:vezora/tested-22k-pyth...   Llama   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/TheBloke/bagel-dpo-34b-v0.2-AWQ 

Bagel DPO 34B V0.2 AWQ Parameters and Internals

Model Type 
yi
Additional Notes 
The model includes multiple format prompt conversions per instruction for enhanced generalization.
Training Details 
Data Sources:
ai2_arc, unalignment/spicy-3.1, codeparrot/apps, facebook/belebele, boolq, jondurbin/cinematika-v0.1, drop, lmsys/lmsys-chat-1m, TIGER-Lab/MathInstruct, cais/mmlu, Muennighoff/natural-instructions, openbookqa, piqa, Vezora/Tested-22k-Python-Alpaca, cakiki/rosetta-code, Open-Orca/SlimOrca, spider, squad_v2, migtissera/Synthia-v1.3, datasets/winogrande, nvidia/HelpSteer, Intel/orca_dpo_pairs, unalignment/toxic-dpo-v0.1, jondurbin/truthy-dpo-v0.1, allenai/ultrafeedback_binarized_cleaned, Squish42/bluemoon-fandom-1-1-rp-cleaned, LDJnr/Capybara, JULIELab/EmoBank, kingbri/PIPPA-shareGPT
Methodology:
Experimental fine-tune of yi-34b-200k using bagel. Includes various DPO datasets including toxic DPO for less censorship.
Input Output 
Input Format:
Prompt templates used: Alpaca (modified), Vicuna, Chat-ML (modified), Llama-2 chat.
LLM NameBagel DPO 34B V0.2 AWQ
Repository 🤗https://huggingface.co/TheBloke/bagel-dpo-34b-v0.2-AWQ 
Model NameBagel DPO 34B v0.2
Model CreatorJon Durbin
Base Model(s)  jondurbin/bagel-dpo-34b-v0.2   jondurbin/bagel-dpo-34b-v0.2
Model Size34b
Required VRAM19.3 GB
Updated2026-07-30
MaintainerTheBloke
Model Typellama
Model Files  10.0 GB: 1-of-2   9.3 GB: 2-of-2
AWQ QuantizationYes
Quantization Typeawq
Model ArchitectureLlamaForCausalLM
Licenseother
Context Length200000
Model Max Length200000
Transformers Version4.37.0.dev0
Tokenizer ClassLlamaTokenizer
Padding Token<unk>
Vocabulary Size64000
Torch Data Typefloat16

Best Alternatives to Bagel DPO 34B V0.2 AWQ

Best Alternatives
Context / RAM
Downloads
Likes
Opus V1 34B AWQ195K / 19.2 GB31
Smaug 34B V0.1 AWQ195K / 19.2 GB32
Yi 34B 200K RPMerge AWQ195K / 19.2 GB51
Tess 34B V1.5B AWQ195K / 19.3 GB73
...34B 200K DARE Megamerge V8 AWQ195K / 19.3 GB102
Deepmoney 34B 200K Base AWQ195K / 19.3 GB252
...ey 34B 200K Chat Evaluator AWQ195K / 19.3 GB55
Nous Capybara Limarpv3 34B AWQ195K / 19.3 GB01
Nontoxic Bagel 34B V0.2 AWQ195K / 19.3 GB112
Bagel 34B V0.2 AWQ195K / 19.3 GB102
Note: green Score (e.g. "73.2") means that the model is better than TheBloke/bagel-dpo-34b-v0.2-AWQ.