What are the hardware requirements for Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2?

Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2 requires approximately 4.2 GB of VRAM and supports a context window of 4K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Where can I download or evaluate Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2?

Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2 by Zoyd — VRAM 4.2GB, 4K context

Name: Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2
Author: Zoyd

Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2 is an open-source language model by Zoyd. Features: LLM, VRAM: 4.2GB, Context: 4K, License: mit, Quantized, Instruction-Based, LLM Explorer Score: 0.13.

Code Conversational Custom code Endpoints compatible Exl2 Instruct Multilingual Phi3 Quantized Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/Zoyd/failspy_Phi-3-medium-4k-instruct-abliterated-v3-2_2bpw_exl2

Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2 Parameters and Internals

Model Type

text-generation

Additional Notes

This model has had certain weights manipulated to 'inhibit' the model's ability to express refusal. It is tuned in all other respects the same as the original 70B instruct model was, just with the strongest refusal directions orthogonalized out.

LLM Name	Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2
Repository 🤗	https://huggingface.co/Zoyd/failspy_Phi-3-medium-4k-instruct-abliterated-v3-2_2bpw_exl2
Required VRAM	4.2 GB
Updated	2026-08-01
Maintainer	Zoyd
Model Type	phi3
Instruction-Based	Yes
Model Files	4.2 GB
Quantization Type	exl2
Model Architecture	Phi3ForCausalLM
License	mit
Context Length	4096
Model Max Length	4096
Transformers Version	4.42.0.dev0
Tokenizer Class	LlamaTokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	32064
Torch Data Type	bfloat16

Rank the Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2 capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

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

Best Alternatives to Failspy Phi 3 Medium 4K Instruct Abliterated V3 2 2bpw EXL2

Best Alternatives	Context / RAM	Downloads	Likes
...m 128K Instruct 8.0bpw H8 EXL2	128K / 13.4 GB	7	4
...m 128K Instruct 6.0bpw H6 EXL2	128K / 10.7 GB	7	3
...dium 128K Instruct 8 0bpw EXL2	128K / 13.4 GB	4	1
...m 128K Instruct 3.0bpw H6 EXL2	128K / 5.6 GB	5	0
...m 128K Instruct 5.0bpw H6 EXL2	128K / 8.9 GB	5	0
...28K Instruct Ov Fp16 Int4 Asym	128K / 2.5 GB	9	0
...128K Instruct HQQ 2bit Smashed	128K / 1.4 GB	8	0
...128K Instruct HQQ 4bit Smashed	128K / 2.9 GB	7	0
NuExtract Bpw6 EXL2	4K / 3 GB	4	1
...Mini 4K Geminified 3 0bpw EXL2	4K / 1.6 GB	6	0

Note: green Score (e.g. "73.2") means that the model is better than Zoyd/failspy_Phi-3-medium-4k-instruct-abliterated-v3-2_2bpw_exl2.