wwewtech/russian-it-community-lora

🦁 Russian IT Community LoRA Model Zoo (55 Foundation Models)

Official Hugging Face Hub Repository: wwewtech/russian-it-community-lora
Total Pre-trained Adapters: 55 Domain Adapters
Training Corpus: 171.5k Curated SFT Multi-Turn Dialogues (Russian IT Community Corpus)
Hardware Target: Optimized for NVIDIA GeForce RTX 3060 (12GB VRAM) & consumer GPUs

---

⚡ Quick Start: 3-Line Inference

import torch
from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "Qwen/Qwen2.5-1.5B-Instruct"
adapter_id = "wwewtech/russian-it-community-lora"
subfolder = "qwen2.5_1.5b_instruct"  # Choose from 55 adapters below

tokenizer = AutoTokenizer.from_pretrained(model_id)
base_model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")
model = PeftModel.from_pretrained(base_model, adapter_id, subfolder=subfolder)

inputs = tokenizer("<|user|>\nКак настроить репликацию PostgreSQL в Kubernetes?<|assistant|>\n", return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

---

📚 Full Catalog of 55 Pre-Trained Adapters

#	Adapter Subfolder	Base Model	Architecture Family	Hugging Face Subfolder Link
01	bloom_1b7	bigscience/bloom-1b7	BigScience BLOOM	bloom_1b7
02	bloom_560m	bigscience/bloom-560m	BigScience BLOOM	bloom_560m
03	deepseek_coder_1.3b_instruct	deepseek-ai/deepseek-coder-1.3b-instruct	DeepSeek	deepseek_coder_1.3b_instruct
04	deepseek_r1_distill_qwen_1.5b	deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B	DeepSeek	deepseek_r1_distill_qwen_1.5b
05	falcon3_1b_instruct	tiiuae/Falcon3-1B-Instruct	TII Falcon	falcon3_1b_instruct
06	falcon3_3b_instruct	tiiuae/Falcon3-3B-Instruct	TII Falcon	falcon3_3b_instruct
07	gemma_2_2b_it	unsloth/gemma-2-2b-it	Open Weights	gemma_2_2b_it
08	gpt2_large	openai-community/gpt2-large	Generative GPT-2 / RuGPT	gpt2_large
09	gpt2_medium	openai-community/gpt2-medium	Generative GPT-2 / RuGPT	gpt2_medium
10	granite_3b_code_instruct	ibm-granite/granite-3b-code-instruct	Open Weights	granite_3b_code_instruct
11	heavyweight_deepseek_r1_7b	unsloth/DeepSeek-R1-Distill-Qwen-7B-bnb-4bit	DeepSeek	heavyweight_deepseek_r1_7b
12	heavyweight_llama3.1_8b	unsloth/Meta-Llama-3.1-8B-Instruct-bnb-4bit	Meta LLaMA	heavyweight_llama3.1_8b
13	heavyweight_qwen2.5_coder_7b	unsloth/Qwen2.5-Coder-7B-Instruct-bnb-4bit	Qwen 2.5 Coder	heavyweight_qwen2.5_coder_7b
14	llama_3.2_1b_instruct	unsloth/Llama-3.2-1B-Instruct	Meta LLaMA	llama_3.2_1b_instruct
15	llama_3.2_3b_instruct	unsloth/Llama-3.2-3B-Instruct	Meta LLaMA	llama_3.2_3b_instruct
16	minicpm_2b_dpo	openbmb/MiniCPM-2B-dpo-bf16	Open Weights	minicpm_2b_dpo
17	opt_1.3b	facebook/opt-1.3b	Meta OPT	opt_1.3b
18	opt_125m	facebook/opt-125m	Meta OPT	opt_125m
19	opt_2.7b	facebook/opt-2.7b	Meta OPT	opt_2.7b
20	opt_350m	facebook/opt-350m	Meta OPT	opt_350m
21	phi_1_5	microsoft/phi-1_5	Microsoft Phi	phi_1_5
22	phi_2	microsoft/phi-2	Microsoft Phi	phi_2
23	phi_3.5_mini_instruct	microsoft/Phi-3.5-mini-instruct	Microsoft Phi	phi_3.5_mini_instruct
24	phi_3_mini_4k_instruct	microsoft/Phi-3-mini-4k-instruct	Microsoft Phi	phi_3_mini_4k_instruct
25	pythia_1.4b	EleutherAI/pythia-1.4b-deduped	EleutherAI Pythia	pythia_1.4b
26	pythia_2.8b	EleutherAI/pythia-2.8b-deduped	EleutherAI Pythia	pythia_2.8b
27	pythia_410m	EleutherAI/pythia-410m-deduped	EleutherAI Pythia	pythia_410m
28	pythia_70m	EleutherAI/pythia-70m-deduped	EleutherAI Pythia	pythia_70m
29	qwen1.5_0.5b_chat	Qwen/Qwen1.5-0.5B-Chat	Qwen 1.5	qwen1.5_0.5b_chat
30	qwen1.5_1.8b_chat	Qwen/Qwen1.5-1.8B-Chat	Qwen 1.5	qwen1.5_1.8b_chat
31	qwen2.5_0.5b_instruct	Qwen/Qwen2.5-0.5B-Instruct	Qwen 2.5	qwen2.5_0.5b_instruct
32	qwen2.5_1.5b_instruct	Qwen/Qwen2.5-1.5B-Instruct	Qwen 2.5	qwen2.5_1.5b_instruct
33	qwen2.5_3b_instruct	Qwen/Qwen2.5-3B-Instruct	Qwen 2.5	qwen2.5_3b_instruct
34	qwen2.5_coder_0.5b_instruct	Qwen/Qwen2.5-Coder-0.5B-Instruct	Qwen 2.5 Coder	qwen2.5_coder_0.5b_instruct
35	qwen2.5_coder_1.5b_instruct	Qwen/Qwen2.5-Coder-1.5B-Instruct	Qwen 2.5 Coder	qwen2.5_coder_1.5b_instruct
36	qwen2.5_coder_3b_instruct	Qwen/Qwen2.5-Coder-3B-Instruct	Qwen 2.5 Coder	qwen2.5_coder_3b_instruct
37	qwen2.5_math_1.5b_instruct	Qwen/Qwen2.5-Math-1.5B-Instruct	Qwen 2.5	qwen2.5_math_1.5b_instruct
38	qwen2_0.5b_instruct	Qwen/Qwen2-0.5B-Instruct	Qwen 2	qwen2_0.5b_instruct
39	qwen2_1.5b_instruct	Qwen/Qwen2-1.5B-Instruct	Qwen 2	qwen2_1.5b_instruct
40	rugpt3_large	ai-forever/rugpt3large_based_on_gpt2	Generative GPT-2 / RuGPT	rugpt3_large
41	rugpt3_medium	ai-forever/rugpt3medium_based_on_gpt2	Generative GPT-2 / RuGPT	rugpt3_medium
42	rugpt3_small	ai-forever/rugpt3small_based_on_gpt2	Generative GPT-2 / RuGPT	rugpt3_small
43	russian_it_lora	Qwen/Qwen2.5-0.5B-Instruct	Open Weights	russian_it_lora
44	smollm2_1.7b_instruct	HuggingFaceTB/SmolLM2-1.7B-Instruct	SmolLM2	smollm2_1.7b_instruct
45	smollm2_135m_instruct	HuggingFaceTB/SmolLM2-135M-Instruct	SmolLM2	smollm2_135m_instruct
46	smollm2_360m_instruct	HuggingFaceTB/SmolLM2-360M-Instruct	SmolLM2	smollm2_360m_instruct
47	smollm_1.7b_instruct	HuggingFaceTB/SmolLM-1.7B-Instruct	SmolLM	smollm_1.7b_instruct
48	smollm_135m_instruct	HuggingFaceTB/SmolLM-135M-Instruct	SmolLM	smollm_135m_instruct
49	smollm_360m_instruct	HuggingFaceTB/SmolLM-360M-Instruct	SmolLM	smollm_360m_instruct
50	stablelm_2_1_6b_chat	stabilityai/stablelm-2-1_6b-chat	Stability AI StableLM	stablelm_2_1_6b_chat
51	stablelm_2_zephyr_1_6b	stabilityai/stablelm-2-zephyr-1_6b	Stability AI StableLM	stablelm_2_zephyr_1_6b
52	tinyllama_1.1b_chat	TinyLlama/TinyLlama-1.1B-Chat-v1.0	Open Weights	tinyllama_1.1b_chat
53	vikhr_llama_3.2_1b	Vikhrmodels/Vikhr-Llama-3.2-1B-instruct	Meta LLaMA	vikhr_llama_3.2_1b
54	vikhr_qwen_2.5_0.5b	Vikhrmodels/Vikhr-Qwen-2.5-0.5B-Instruct	Vikhr Russian NLP	vikhr_qwen_2.5_0.5b
55	vikhr_qwen_2.5_1.5b	Vikhrmodels/Vikhr-Qwen-2.5-1.5B-Instruct	Vikhr Russian NLP	vikhr_qwen_2.5_1.5b