Text Generation
Transformers
Safetensors
English
lowonmind
tiny-lm
pretrained-from-scratch
scaling-limits
custom_code
Instructions to use DedeProGames/LowOnMind-5M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use DedeProGames/LowOnMind-5M with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="DedeProGames/LowOnMind-5M", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("DedeProGames/LowOnMind-5M", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use DedeProGames/LowOnMind-5M with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "DedeProGames/LowOnMind-5M" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "DedeProGames/LowOnMind-5M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/DedeProGames/LowOnMind-5M
- SGLang
How to use DedeProGames/LowOnMind-5M 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 "DedeProGames/LowOnMind-5M" \ --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": "DedeProGames/LowOnMind-5M", "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 "DedeProGames/LowOnMind-5M" \ --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": "DedeProGames/LowOnMind-5M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use DedeProGames/LowOnMind-5M with Docker Model Runner:
docker model run hf.co/DedeProGames/LowOnMind-5M
| { | |
| "data": "fineweb_edu_200M_v1024.uint16.bin", | |
| "base": "LowOnMind-5M", | |
| "derived_from": "DedeProGames/LowOnMind-300k, DedeProGames/LowOnMind-1M", | |
| "dataset": "HuggingFaceFW/fineweb-edu:sample-10BT", | |
| "tokenizer": "byte-level BPE 1024", | |
| "params": 4920384, | |
| "seq_len": 512, | |
| "batch_size": 64, | |
| "grad_accum": 1, | |
| "max_steps": 6103, | |
| "total_tokens": 199983104, | |
| "lr": 0.0012, | |
| "min_lr": 0.00012, | |
| "warmup_steps": 250, | |
| "weight_decay": 0.1, | |
| "grad_clip": 1.0, | |
| "val_tokens": 2000000, | |
| "eval_every": 500, | |
| "eval_batches": 40, | |
| "device": "cuda", | |
| "dtype": "float16", | |
| "compile": true, | |
| "seed": 1337, | |
| "train_minutes": 27.0, | |
| "final_val_loss": 2.5828, | |
| "final_val_ppl": 13.235, | |
| "bits_per_char": 1.5856, | |
| "chars_per_token": 2.35, | |
| "reused_tokenizer_from": "DedeProGames/LowOnMind-300k", | |
| "embedding_share": 0.04, | |
| "baseline_300k": { | |
| "name": "LowOnMind-300k", | |
| "params": 296960, | |
| "val_loss": 3.2982, | |
| "val_ppl": 27.06, | |
| "bpc": 2.03, | |
| "bench_elo": 833, | |
| "bench_acc": 0.266, | |
| "bench_lang": 0.46, | |
| "real_word": null | |
| }, | |
| "baseline_1m": { | |
| "name": "LowOnMind-1M", | |
| "params": 985152, | |
| "val_loss": 2.9908, | |
| "val_ppl": 19.9, | |
| "bpc": 1.836, | |
| "bench_elo": 843, | |
| "bench_acc": 0.289, | |
| "bench_lang": 0.52, | |
| "real_word": 0.98 | |
| }, | |
| "lexicon": { | |
| "model_real_word_rate": 0.963, | |
| "corpus_real_word_rate": 0.9839, | |
| "words_scored": 5398, | |
| "samples": 64, | |
| "top_nonwords": [ | |
| [ | |
| "shima", | |
| 13 | |
| ], | |
| [ | |
| "meikimo", | |
| 11 | |
| ], | |
| [ | |
| "amiravicis", | |
| 8 | |
| ], | |
| [ | |
| "illuminator", | |
| 3 | |
| ], | |
| [ | |
| "hymenola", | |
| 3 | |
| ], | |
| [ | |
| "phillipsburg", | |
| 3 | |
| ], | |
| [ | |
| "perleti", | |
| 3 | |
| ], | |
| [ | |
| "almanine", | |
| 2 | |
| ], | |
| [ | |
| "morrisky", | |
| 2 | |
| ], | |
| [ | |
| "anthocyanin", | |
| 2 | |
| ], | |
| [ | |
| "shekarwons", | |
| 2 | |
| ], | |
| [ | |
| "ziko", | |
| 1 | |
| ], | |
| [ | |
| "basiles", | |
| 1 | |
| ], | |
| [ | |
| "gooverton", | |
| 1 | |
| ], | |
| [ | |
| "mcleil", | |
| 1 | |
| ] | |
| ] | |
| }, | |
| "samples": [ | |
| { | |
| "prompt": "The ", | |
| "text": "The Xylococcus cannot sound alterarian and apnea.\nSimilarly, a bacteria may be calcium, and the Xylococcus cannot be cut into the mouth. This cannot be reduced and the muscles can\u2019t be treated by slowing the mouth. Bacteria can also cause problems in the vomiting and sleeping of the bact" | |
| }, | |
| { | |
| "prompt": "Photosynthesis is ", | |
| "text": "Photosynthesis is essential to the human tissues. At the end of the vision, the body can act as a method that makes it easier to understand the brain. What is the brain, it can be the first to use the brain to perform a brain function, but it is important to continue to live. For example, in addition to the brain cells that receive the brain, we have to" | |
| }, | |
| { | |
| "prompt": "In 1969, ", | |
| "text": "In 1969, 86 percent of the people from the Palestine candidates had received a higher rate of jobs, and they were brought to the Civil War.\nPalestinian candidates initiated a free financial and copyright against the War of the Region. The country was the second century bankers, now bankers, and money, within which they were" | |
| }, | |
| { | |
| "prompt": "Students should ", | |
| "text": "Students should identify their strengths, and gain control of their bodies.\nThe results of that study were introduced in the BMCP (September 2018). BMCP (September 2015) and the researchers identified the benefits of drugs in humans, and these studies have been released to the study. The results suggested that drugs that were strong" | |
| } | |
| ], | |
| "history": [ | |
| { | |
| "step": 500, | |
| "val_loss": 3.3853, | |
| "val_ppl": 29.528 | |
| }, | |
| { | |
| "step": 1000, | |
| "val_loss": 3.1327, | |
| "val_ppl": 22.936 | |
| }, | |
| { | |
| "step": 1500, | |
| "val_loss": 2.9676, | |
| "val_ppl": 19.446 | |
| }, | |
| { | |
| "step": 2000, | |
| "val_loss": 2.8651, | |
| "val_ppl": 17.55 | |
| }, | |
| { | |
| "step": 2500, | |
| "val_loss": 2.7969, | |
| "val_ppl": 16.393 | |
| }, | |
| { | |
| "step": 3000, | |
| "val_loss": 2.7501, | |
| "val_ppl": 15.644 | |
| }, | |
| { | |
| "step": 3500, | |
| "val_loss": 2.7143, | |
| "val_ppl": 15.093 | |
| }, | |
| { | |
| "step": 4000, | |
| "val_loss": 2.6789, | |
| "val_ppl": 14.569 | |
| }, | |
| { | |
| "step": 4500, | |
| "val_loss": 2.6498, | |
| "val_ppl": 14.151 | |
| }, | |
| { | |
| "step": 5000, | |
| "val_loss": 2.6273, | |
| "val_ppl": 13.837 | |
| }, | |
| { | |
| "step": 5500, | |
| "val_loss": 2.6074, | |
| "val_ppl": 13.564 | |
| }, | |
| { | |
| "step": 6000, | |
| "val_loss": 2.6067, | |
| "val_ppl": 13.554 | |
| }, | |
| { | |
| "step": 6103, | |
| "val_loss": 2.6016, | |
| "val_ppl": 13.486 | |
| } | |
| ] | |
| } |