GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
pulipakav-1
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
pulipakav-1
Model Tree
Input Modalities
Output Modalities
Supported Functionality
| Property | Value |
|---|---|
| Architecture | GPT-2 (Causal LM) |
| Parameters | 110M |
| Layers | 12 |
| Attention Heads | 12 |
| Hidden Size | 768 |
| Context Length | 1024 |
| Vocab Size | 32,000 |
| Tokenizer | SentencePiece BPE |
The model was trained on a combination of Malayalam corpora:
| Source | Description |
|---|---|
| CC-100 | Web-crawled Malayalam text from CommonCrawl |
| Wikipedia (20231101.ml) | Malayalam Wikipedia dump (November 2023) |
| Samanantar | Malayalam parallel corpus |
| TinyStories (ml) | Cleaned Malayalam short stories |
python
import sentencepiece as spmfrom huggingface_hub import hf_hub_downloadfrom transformers import AutoModelForCausalLMimport torch # Load modelmodel = AutoModelForCausalLM.from_pretrained("pulipakav-1/dravidian-gpt2-malayalam")model.eval() # Load tokenizermodel_path = hf_hub_download( repo_id="pulipakav-1/dravidian-gpt2-malayalam", filename="tokenizer.model")sp = spm.SentencePieceProcessor()sp.Load(model_path) # Generateprompt = "മലയാളം ഭാഷ"token_ids = sp.Encode(prompt)inputs = {"input_ids": torch.tensor([token_ids])} with torch.no_grad(): outputs = model.generate( **inputs, max_new_tokens=50, do_sample=True, temperature=0.7, top_p=0.9, repetition_penalty=1.2, pad_token_id=0, eos_token_id=3, bos_token_id=2 ) print(sp.Decode(outputs[0].tolist()))If you use this model in your research, please cite:
bibtex
@misc{dravidian-gpt2-malayalam, author = {pulipakav-1}, title = {Dravidian GPT-2 Malayalam}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/pulipakav-1/dravidian-gpt2-malayalam}}This model is released under the MIT License.
import sentencepiece as spmfrom huggingface_hub import hf_hub_downloadfrom transformers import AutoModelForCausalLMimport torch # Load modelmodel = AutoModelForCausalLM.from_pretrained("pulipakav-1/dravidian-gpt2-malayalam")model.eval() # Load tokenizermodel_path = hf_hub_download( repo_id="pulipakav-1/dravidian-gpt2-malayalam", filename="tokenizer.model")sp = spm.SentencePieceProcessor()sp.Load(model_path) # Generateprompt = "മലയാളം ഭാഷ"token_ids = sp.Encode(prompt)inputs = {"input_ids": torch.tensor([token_ids])} with torch.no_grad(): outputs = model.generate( **inputs, max_new_tokens=50, do_sample=True, temperature=0.7, top_p=0.9, repetition_penalty=1.2, pad_token_id=0, eos_token_id=3, bos_token_id=2 ) print(sp.Decode(outputs[0].tolist()))@misc{dravidian-gpt2-malayalam, author = {pulipakav-1}, title = {Dravidian GPT-2 Malayalam}, year = {2026}, publisher = {Hugging Face}, url = {https://huggingface.co/pulipakav-1/dravidian-gpt2-malayalam}}