@bigdatai
Новый канал "Big Data AI" - это идеальное место для всех, кто интересуется современными технологиями в области искусственного интеллекта, машинного обучения и анализа данных. Канал предлагает уникальную возможность погружения в мир больших данных и исследований в области искусственного интеллекта.
@haarrp, администратор канала, делится самой актуальной и интересной информацией о вопросах, задаваемых на собеседованиях по Machine Learning, Data Science, Deep Learning и Нейронным сетям. Вы узнаете о последних трендах в этой области, а также сможете задать вопросы и обсудить интересующие темы с опытными специалистами.
Помимо этого, вы найдете ссылки на другие каналы, посвященные анализу данных, машинному обучению и большим данным. Каналы @data_analysis_ml, @ai_machinelearning_big_data и @itchannels_telegram помогут вам расширить свои знания и навыки в сфере программирования и аналитики данных.
Присоединяйтесь к каналу "Big Data AI" прямо сейчас и станьте частью сообщества людей, увлеченных изучением и применением передовых технологий в области искусственного интеллекта и анализа данных.
17 Feb, 16:26
16 Feb, 07:53
$ pip install podcastfy
15 Feb, 08:04
15 Feb, 06:00
14 Feb, 06:11
13 Feb, 11:03
13 Feb, 09:59
12 Feb, 05:51
11 Feb, 05:54
10 Feb, 09:22
07 Feb, 07:42
llm install llm-smollm2
06 Feb, 12:38
05 Feb, 15:54
04 Feb, 18:31
04 Feb, 16:26
04 Feb, 14:03
04 Feb, 11:00
04 Feb, 09:01
02 Feb, 12:05
01 Feb, 11:00
31 Jan, 09:14
29 Jan, 11:42
28 Jan, 15:23
27 Jan, 17:30
27 Jan, 17:13
27 Jan, 14:25
26 Jan, 18:08
25 Jan, 11:02
24 Jan, 16:57
GatedDeltaNet-H1 и GatedDeltaNet-H2
дали еще более высокие результаты, используя комбинации Gated DeltaNet + SWA и Mamba2 + Gated DeltaNet + SWA соответственно.23 Jan, 06:27
23 Jan, 05:33
git clone https://github.com/DepthAnything/Video-Depth-Anything
cd Video-Depth-Anything
pip install -r requirements.txt
21 Jan, 11:01
20 Jan, 10:28
pip install smolagents
from smolagents import CodeAgent, DuckDuckGoSearchTool, HfApiModel
agent = CodeAgent(tools=[DuckDuckGoSearchTool()], model=HfApiModel())
agent.run("How many seconds would it take for a leopard at full speed to run through Pont des Arts?")
19 Jan, 15:54
19 Jan, 14:42
Python, Java, C++, JavaScript, C#
и другие!
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModel
# Each query needs to be accompanied by an corresponding instruction describing the task.
query_instruction_example = "Given Code or Text, retrieval relevant content"
queries = [
"how to implement quick sort in Python?"
]
# No instruction needed for retrieval passages
passages = [
"def quick_sort(arr):\n if len(arr) <= 1:\n return arr\n pivot = arr[len(arr) // 2]\n left = [x for x in arr if x < pivot]\n middle = [x for x in arr if x == pivot]\n right = [x for x in arr if x > pivot]\n return quick_sort(left) + middle + quick_sort(right)",
"def bubble_sort(arr):\n n = len(arr)\n for i in range(n):\n for j in range(0, n-i-1):\n if arr[j] > arr[j+1]:\n arr[j], arr[j+1] = arr[j+1], arr[j]\n return arr"
]
# load model with tokenizer
model = AutoModel.from_pretrained('Salesforce/SFR-Embedding-Code-2B_R', trust_remote_code=True)
# get the embeddings
max_length = 32768
query_embeddings = model.encode_queries(queries, instruction=query_instruction_example, max_length=max_length)
passage_embeddings = model.encode_corpus(passages, max_length=max_length)
# normalize embeddings
query_embeddings = F.normalize(query_embeddings, p=2, dim=1)
passage_embeddings = F.normalize(passage_embeddings, p=2, dim=1)
scores = (query_embeddings @ passage_embeddings.T) * 100
print(scores.tolist())
18 Jan, 13:04
18 Jan, 10:53
18 Jan, 10:02
16 Jan, 14:02
14 Jan, 10:10
14 Jan, 08:03
08 Jan, 09:35
06 Jan, 10:03
04 Jan, 13:04
02 Jan, 12:05
30 Dec, 15:26
29 Dec, 16:42
29 Dec, 14:42
29 Dec, 10:04
28 Dec, 13:00
27 Dec, 11:27
26 Dec, 16:15
# Clone repo
git clone https://github.com/Johanan528/DepthLab.git
cd DepthLab
# Create conda env
conda env create -f environment.yaml
conda activate DepthLab
# Run inference
cd scripts
bash infer.sh
26 Dec, 09:00
26 Dec, 07:01
25 Dec, 15:21
24 Dec, 12:50
23 Dec, 10:50
22 Dec, 16:40
22 Dec, 13:01
21 Dec, 12:00
21 Dec, 11:32
20 Dec, 11:00
20 Dec, 09:01
19 Dec, 16:09
25 Nov, 13:01
23 Nov, 12:01
22 Nov, 14:47
22 Nov, 13:32
22 Nov, 10:01
22 Nov, 08:01
19 Nov, 14:00
17 Nov, 13:00
16 Nov, 11:00
16 Nov, 09:36
15 Nov, 16:00
15 Nov, 08:56
13 Nov, 09:20
12 Nov, 10:03
12 Nov, 08:01
11 Nov, 17:00
11 Nov, 15:02
11 Nov, 12:08
11 Nov, 10:08
SELECT *
FROM subscribers
WHERE channel_name = 'bigdatai'
AND technical_skills IN ('SQL', 'Airflow', 'MapReduce', 'DataLens')
AND data_driven_approach = true
AND analytical_mindset = true
AND years_of_experience >= 2
AND fit = true;
11 Nov, 07:11
10 Nov, 15:01
09 Nov, 14:00
08 Nov, 10:01
08 Nov, 08:01
07 Nov, 15:32
07 Nov, 14:00
07 Nov, 11:58
from transformers import AutoModelForCausalLM, AutoTokenizer
checkpoint = "HuggingFaceTB/SmolLM2-1.7B"
device = "cuda" # for GPU usage or "cpu" for CPU usage
tokenizer = AutoTokenizer.from_pretrained(checkpoint)
model = AutoModelForCausalLM.from_pretrained(checkpoint).to(device)
inputs = tokenizer.encode("Gravity is", return_tensors="pt").to(device)
outputs = model.generate(inputs)
print(tokenizer.decode(outputs[0]))
07 Nov, 06:29
07 Nov, 04:28
01 Nov, 14:00
31 Oct, 11:01
31 Oct, 09:01
30 Oct, 15:01
30 Oct, 10:53
29 Oct, 14:01
28 Oct, 13:00
27 Oct, 12:00
27 Oct, 10:00
26 Oct, 13:28
# Clone repo
git clone https://github.com/genmoai/models
cd models
# Install using uv
pip install uv
uv venv .venv
source .venv/bin/activate
uv pip install -e .
# Inference with Gradio UI
python3 -m mochi_preview.gradio_ui --model_dir "<path_to_model_directory>"
# Inference with CLI
python3 -m mochi_preview.infer --prompt "%prompt%" --seed 1710977262 --cfg_scale 4.5 --model_dir "<path_to_model_directory>"
26 Oct, 10:00
26 Oct, 08:02
25 Oct, 08:26
24 Oct, 13:19
24 Oct, 12:02
22 Oct, 19:35
22 Oct, 17:46
22 Oct, 15:49
21 Oct, 14:01
20 Oct, 12:01
20 Oct, 10:19
use_mamba_kernels=False
при загрузке модели с помощью AutoModelForCausalLM.from_pretrained
.# Clone repo
git clone https://github.com/Zyphra/transformers_zamba2.git
cd transformers_zamba2
# Install the repository & accelerate:
pip install -e .
pip install accelerate
# Inference:
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch
tokenizer = AutoTokenizer.from_pretrained("Zyphra/Zamba2-2.7B-instruct")
model = AutoModelForCausalLM.from_pretrained("Zyphra/Zamba2-2.7B-instruct", device_map="cuda", torch_dtype=torch.bfloat16)
user_turn_1 = "user_prompt1."
assistant_turn_1 = "assistant_prompt."
user_turn_2 = "user_prompt2."
sample = [{'role': 'user', 'content': user_turn_1}, {'role': 'assistant', 'content': assistant_turn_1}, {'role': 'user', 'content': user_turn_2}]
chat_sample = tokenizer.apply_chat_template(sample, tokenize=False)
input_ids = tokenizer(chat_sample, return_tensors='pt', add_special_tokens=False).to("cuda")
outputs = model.generate(**input_ids, max_new_tokens=150, return_dict_in_generate=False, output_scores=False, use_cache=True, num_beams=1, do_sample=False)
print((tokenizer.decode(outputs[0])))
18 Oct, 16:00
16 Oct, 17:09
16 Oct, 16:01
16 Oct, 14:56
# Install miniGRU-pytorch
pip install minGRU-pytorch
# Usage
import torch
from minGRU_pytorch import minGRU
min_gru = minGRU(512)
x = torch.randn(2, 1024, 512)
out = min_gru(x)
assert x.shape == out.shape
# Sanity check
import torch
from minGRU_pytorch import minGRU
min_gru = minGRU(dim = 512, expansion_factor = 1.5)
x = torch.randn(1, 2048, 512)
# parallel
parallel_out = min_gru(x)[:, -1:]
# sequential
prev_hidden = None
for token in x.unbind(dim = 1):
sequential_out, prev_hidden = min_gru(token[:, None, :], prev_hidden, return_next_prev_hidden = True)
assert torch.allclose(parallel_out, sequential_out, atol = 1e-4)
15 Oct, 11:30
15 Oct, 09:23
14 Oct, 17:00
14 Oct, 15:00
14 Oct, 10:00
14 Oct, 08:01
13 Oct, 10:01
13 Oct, 08:19
11 Oct, 16:03
11 Oct, 07:03
10 Oct, 09:07
08 Oct, 18:34
08 Oct, 16:41