Quantization made by Richard Erkhov.

dictalm2.0 - GGUF

Model creator: https://huggingface.co/dicta-il/
Original model: https://huggingface.co/dicta-il/dictalm2.0/

Name	Quant method	Size
dictalm2.0.Q2_K.gguf	Q2_K	2.54GB
dictalm2.0.IQ3_XS.gguf	IQ3_XS	2.82GB
dictalm2.0.IQ3_S.gguf	IQ3_S	2.97GB
dictalm2.0.Q3_K_S.gguf	Q3_K_S	2.95GB
dictalm2.0.IQ3_M.gguf	IQ3_M	3.06GB
dictalm2.0.Q3_K.gguf	Q3_K	3.28GB
dictalm2.0.Q3_K_M.gguf	Q3_K_M	3.28GB
dictalm2.0.Q3_K_L.gguf	Q3_K_L	3.57GB
dictalm2.0.IQ4_XS.gguf	IQ4_XS	3.68GB
dictalm2.0.Q4_0.gguf	Q4_0	3.83GB
dictalm2.0.IQ4_NL.gguf	IQ4_NL	3.88GB
dictalm2.0.Q4_K_S.gguf	Q4_K_S	3.86GB
dictalm2.0.Q4_K.gguf	Q4_K	4.07GB
dictalm2.0.Q4_K_M.gguf	Q4_K_M	4.07GB
dictalm2.0.Q4_1.gguf	Q4_1	4.25GB
dictalm2.0.Q5_0.gguf	Q5_0	4.66GB
dictalm2.0.Q5_K_S.gguf	Q5_K_S	4.66GB
dictalm2.0.Q5_K.gguf	Q5_K	4.79GB
dictalm2.0.Q5_K_M.gguf	Q5_K_M	4.79GB
dictalm2.0.Q5_1.gguf	Q5_1	5.08GB
dictalm2.0.Q6_K.gguf	Q6_K	5.54GB
dictalm2.0.Q8_0.gguf	Q8_0	7.18GB

Original model description:

license: apache-2.0 pipeline_tag: text-generation language: - en - he tags: - pretrained inference: parameters: temperature: 0.7

Adapting LLMs to Hebrew: Unveiling DictaLM 2.0 with Enhanced Vocabulary and Instruction Capabilities

The DictaLM-2.0 Large Language Model (LLM) is a pretrained generative text model with 7 billion parameters trained to specialize in Hebrew text.

For full details of this model please read our release blog post or the technical report.

This is the full-precision base model. You can view and access the full collection of base/instruct unquantized/quantized versions of DictaLM-2.0 here.

Example Code

from transformers import pipeline
import torch

# This loads the model onto the GPU in bfloat16 precision
model = pipeline('text-generation', 'dicta-il/dictalm2.0', torch_dtype=torch.bfloat16, device_map='cuda')

# Sample few shot examples
prompt = """
עבר: הלכתי
עתיד: אלך

עבר: שמרתי
עתיד: אשמור

עבר: שמעתי
עתיד: אשמע

עבר: הבנתי
עתיד:
"""

print(model(prompt.strip(), do_sample=False, max_new_tokens=8, stop_sequence='\n'))
# [{'generated_text': 'עבר: הלכתי\nעתיד: אלך\n\nעבר: שמרתי\nעתיד: אשמור\n\nעבר: שמעתי\nעתיד: אשמע\n\nעבר: הבנתי\nעתיד: אבין\n\n'}]

Example Code - 4-Bit

There are already pre-quantized 4-bit models using the GPTQ and AWQ methods available for use: DictaLM-2.0-AWQ and DictaLM-2.0-GPTQ.

For dynamic quantization on the go, here is sample code which loads the model onto the GPU using the bitsandbytes package, requiring :

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained('dicta-il/dictalm2.0', torch_dtype=torch.bfloat16, device_map='cuda', load_in_4bit=True)
tokenizer = AutoTokenizer.from_pretrained('dicta-il/dictalm2.0')

prompt = """
עבר: הלכתי
עתיד: אלך

עבר: שמרתי
עתיד: אשמור

עבר: שמעתי
עתיד: אשמע

עבר: הבנתי
עתיד:
"""

encoded = tokenizer(prompt.strip(), return_tensors='pt').to(model.device)
print(tokenizer.batch_decode(model.generate(**encoded, do_sample=False, max_new_tokens=4)))
# ['<s> עבר: הלכתי\nעתיד: אלך\n\nעבר: שמרתי\nעתיד: אשמור\n\nעבר: שמעתי\nעתיד: אשמע\n\nעבר: הבנתי\nעתיד: אבין\n\n']

Model Architecture

DictaLM-2.0 is based on the Mistral-7B-v0.1 model with the following changes:

An extended tokenizer with 1,000 injected tokens specifically for Hebrew, increasing the compression rate from 5.78 tokens/word to 2.76 tokens/word.
Continued pretraining on over 190B tokens of naturally occuring text, 50% Hebrew and 50% English.

Notice

DictaLM 2.0 is a pretrained base model and therefore does not have any moderation mechanisms.

Citation

If you use this model, please cite:

@misc{shmidman2024adaptingllmshebrewunveiling,
      title={Adapting LLMs to Hebrew: Unveiling DictaLM 2.0 with Enhanced Vocabulary and Instruction Capabilities}, 
      author={Shaltiel Shmidman and Avi Shmidman and Amir DN Cohen and Moshe Koppel},
      year={2024},
      eprint={2407.07080},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2407.07080}, 
}