What are the hardware requirements for Dictalm2.0?

Dictalm2.0 requires approximately 14.5 GB of VRAM and supports a context window of 32K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Dictalm2.0 and how large is it?

Dictalm2.0 is developed by dicta-il, a model with 7.3b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Dictalm2.0?

Dictalm2.0 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Dictalm2.0 by dicta-il — VRAM 14.5GB, 32K context

Name: Dictalm2.0
Author: dicta-il

Dictalm2.0 is an open-source language model by dicta-il. Features: 7.3b LLM, VRAM: 14.5GB, Context: 32K, License: apache-2.0, LLM Explorer Score: 0.16.

Arxiv:2407.07080 Deploy:azure En Endpoints compatible He Mistral Pretrained Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/dicta-il/dictalm2.0

Dictalm2.0 Parameters and Internals

Model Type

text generation

Use Cases

Areas:

text generation

Limitations:

no moderation mechanisms

Additional Notes

DictaLM-2.0 is a pretrained base model specifically designed to specialize in Hebrew text.

Supported Languages

en (proficient), he (proficient)

Training Details

Data Volume:

over 190B tokens

Methodology:

extended tokenizer with 1,000 injected tokens for Hebrew

Model Architecture:

Mistral-7B-v0.1 with enhancements for the Hebrew language

Input Output

Accepted Modalities:

text

Output Format:

generated text

LLM Name	Dictalm2.0
Repository 🤗	https://huggingface.co/dicta-il/dictalm2.0
Model Size	7.3b
Required VRAM	14.5 GB
Updated	2026-07-21
Maintainer	dicta-il
Model Type	mistral
Model Files	5.0 GB: 1-of-3 5.0 GB: 2-of-3 4.5 GB: 3-of-3
Supported Languages	en he
Model Architecture	MistralForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.37.2
Tokenizer Class	LlamaTokenizer
Vocabulary Size	33152
Torch Data Type	bfloat16

Rank the Dictalm2.0 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

Quantized Models of the Dictalm2.0

Model	Likes	Downloads	VRAM
Dictalm2.0 Instruct GGUF	7	664	4 GB
Dictalm2.0 Instruct AWQ	0	93	4 GB
Dictalm2.0 Instruct GPTQ	0	6	4 GB

Best Alternatives to Dictalm2.0

Best Alternatives	Context / RAM	Downloads	Likes
Dictalm2.0 Instruct	32K / 14.5 GB	2791	25
My Tokenizer 100 10000	32K / 22.3 GB	12	0
My Tokenizer 50 20000	32K / 14.6 GB	7	0
Mis	32K / 14.6 GB	6	0
Dictalm2 It Qa Fine Tune	32K / 14.5 GB	26	7
Misjava Api 060924 V3 Merged	32K / 14.6 GB	5	0
... Fine Tuned Alpaca Gpt4 Hebrew	32K / 14.5 GB	201	0
Dictalm2.0 Instruct Fine Tuned	32K / 14.5 GB	21	0
Quietstar 8 Ahead	32K / 14.5 GB	43	90
Swallow Hermes St V1	4K / 14.6 GB	14	14

Note: green Score (e.g. "73.2") means that the model is better than dicta-il/dictalm2.0.