Text Generation
Transformers
Safetensors
English
llama
small-language-model
from-scratch
sentence-scoring
sentence-ranking
text-scoring
tiny-model
english
34m
causal-lm
language-model
perplexity
grammar
sentence-judge
small-model
pretrained-from-scratch
fineweb-edu
wikipedia
gutenberg
open-data
colab
kaggle
slm
nano-model
lightweight
edge-ai
on-device
cpu
offline
local-ai
base-model
pretraining
language-modeling
nlp
llama-architecture
rope
swiglu
rmsnorm
custom-tokenizer
byte-level-bpe
reranking
sentence-acceptability
fluency-scoring
grammar-checking
education
research
cc-by-sa-data
text-generation-inference
Instructions to use vrishv/Viveka-34M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use vrishv/Viveka-34M with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="vrishv/Viveka-34M")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("vrishv/Viveka-34M") model = AutoModelForCausalLM.from_pretrained("vrishv/Viveka-34M", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use vrishv/Viveka-34M with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "vrishv/Viveka-34M" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "vrishv/Viveka-34M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/vrishv/Viveka-34M
- SGLang
How to use vrishv/Viveka-34M with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "vrishv/Viveka-34M" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "vrishv/Viveka-34M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "vrishv/Viveka-34M" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "vrishv/Viveka-34M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use vrishv/Viveka-34M with Docker Model Runner:
docker model run hf.co/vrishv/Viveka-34M
Commit History
Yesterday
-
Viveka v3: IAM-free retrain, now CC BY-SA 4.0 (perplexity 176.9 -> 137.3)
Verified signatured5bafc2
13 days ago
-
Viveka v2: better sentence judge; CC BY-NC 4.0 (first stage used IAM texts); new card
Verified signature119b9c0 -
Retrained from scratch without IAM; MIT licence; new card
Verified signature5d1702b
17 days ago
-
Viveka-34M: small English sentence-scoring model trained from scratch
Verified signature4fea3cd