LLM EXPLORER 59,516 MODELS INDEXED

TaraV2 M16B by Gragroo

By Gragroo · 5 downloads
HF Repo ↗ Copy URL

TaraV2 M16B is an open-source language model by Gragroo. Features: 3.2b LLM, VRAM: 6.5GB, Context: 128K, License: apache-2.0, Quantized, Merged, LLM Explorer Score: 0.15.

  Merged Model   16bit   Autotrain compatible Base model:finetune:gragroo/ta...   Base model:gragroo/tarav1-m16b   Conversational   En   Endpoints compatible   Llama   Quantized   Region:us   Safetensors   Sharded   Tensorflow   Trl   Unsloth
Model Card on HF 🤗: https://huggingface.co/Gragroo/taraV2-M16B 

TaraV2 M16B Parameters and Internals

LLM NameTaraV2 M16B
Repository 🤗https://huggingface.co/Gragroo/taraV2-M16B 
Base Model(s)  RunpodV1 Merged 16bit   Gragroo/runpodV1-merged_16bit
Merged ModelYes
Model Size3.2b
Required VRAM6.5 GB
Updated2025-09-23
MaintainerGragroo
Model Typellama
Model Files  5.0 GB: 1-of-2   1.5 GB: 2-of-2
Supported Languagesen
Quantization Type16bit
Model ArchitectureLlamaForCausalLM
Licenseapache-2.0
Context Length131072
Model Max Length131072
Transformers Version4.44.2
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|finetune_right_pad_id|>
Vocabulary Size128256
Torch Data Typebfloat16

Best Alternatives to TaraV2 M16B

Best Alternatives
Context / RAM
Downloads
Likes
Q448128K / 6.5 GB120
...uvery Train 0.0.2 Merged 16bit128K / 6.5 GB60
Tc2128K / 6.5 GB50
Rs44128K / 6.5 GB5820
Llama3.2 Resized128K / 6.5 GB50
BG58128K / 6.5 GB80
Xddd128K / 6.5 GB60
ColdBrew Oracle128K / 6.5 GB50
ISLEXITAS WEL SHREE 02 12 24128K / 6.5 GB150
Testing V01128K / 6.5 GB200
Note: green Score (e.g. "73.2") means that the model is better than Gragroo/taraV2-M16B.