LLM EXPLORER 58,873 MODELS INDEXED

Phi 3 Orpo V9.4 by cstr

By cstr · 6 downloads
HF Repo ↗ Copy URL

Phi 3 Orpo V9.4 is an open-source language model by cstr. Features: 3.8b LLM, VRAM: 2.3GB, Context: 4K, License: apache-2.0, LLM Explorer Score: 0.12.

  4-bit Base model:cstr/phi-3-orpo-v8 ... Base model:quantized:cstr/phi-...   Bitsandbytes   Conversational   En   Endpoints compatible   Llama   Orpo   Region:us   Safetensors   Trl   Unsloth
Model Card on HF 🤗: https://huggingface.co/cstr/phi-3-orpo-v9_4 

Phi 3 Orpo V9.4 Parameters and Internals

Model Type 
text-generation-inference, transformers, unsloth, llama, trl, orpo
Additional Notes 
The model was trained 2x faster using Unsloth.
Training Details 
Methodology:
Unsloth and Huggingface's TRL library
LLM NamePhi 3 Orpo V9.4
Repository 🤗https://huggingface.co/cstr/phi-3-orpo-v9_4 
Base Model(s)  Phi 3 Orpo V8.16   cstr/phi-3-orpo-v8_16
Model Size3.8b
Required VRAM2.3 GB
Updated2026-08-03
Maintainercstr
Model Typellama
Model Files  2.3 GB
Supported Languagesen
Model ArchitectureLlamaForCausalLM
Licenseapache-2.0
Context Length4096
Model Max Length4096
Transformers Version4.40.0
Tokenizer ClassLlamaTokenizer
Padding Token<|endoftext|>
Vocabulary Size32064
Torch Data Typefloat16

Best Alternatives to Phi 3 Orpo V9.4

Best Alternatives
Context / RAM
Downloads
Likes
Tara 3.8B V1.1146K / 7.6 GB151
Tara 3.8B146K / 7.6 GB100
Veena128K / 7.6 GB4252235
Indic Tts Lora Bengali Stage 3128K / 9.5 GB50
Indic Tts Lora Tamil Stage 2128K / 9.5 GB80
Indic Tts Lora Tamil Stage 3128K / 9.5 GB50
Indic Tts Lora Hindi Stage 3128K / 9.5 GB50
Indic Tts Lora Hindi Stage 2128K / 9.5 GB50
...hi3 Unlearning General Methode128K / 7.6 GB3550
Superthoughts Mini V1.3.8B128K / 7.6 GB781
Note: green Score (e.g. "73.2") means that the model is better than cstr/phi-3-orpo-v9_4.