GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
Xangel0s
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
Xangel0s
Model Tree
unsloth/Qwen2.5-Coder-1.5B-Instruct-bnb-4bit
Input Modalities
Output Modalities
Supported Functionality
/v1/chat/completions).| Prueba | Objetivo | Resultado |
|---|---|---|
| Algoritmos Complejos | LRU Cache O(1) con Doubly Linked List | ✅ 100% Correcto |
| Análisis de Concurrencia | Detección de Deadlock & Prevención | ✅ 100% Detectado |
| Auditoría de Alucinaciones | Detección de librería y sintaxis falsa | ✅ 0.0% Alucinación |
| Razonamiento Big-O | Demostración matemática O(N log N) | ✅ 100% Riguroso |
python
import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer model_id = "Xangel0s/OzyTalk-1.58b"base_id = "unsloth/Qwen2.5-Coder-1.5B-Instruct-bnb-4bit" tokenizer = AutoTokenizer.from_pretrained(model_id)base_model = AutoModelForCausalLM.from_pretrained( base_id, torch_dtype=torch.bfloat16, device_map="auto")model = PeftModel.from_pretrained(base_model, model_id)model.eval() prompt = "Explica la diferencia entre procesos e hilos y cuando usar cada uno."messages = [ {"role": "system", "content": "Eres OzyTalk, un asistente de programacion amigable y preciso."}, {"role": "user", "content": prompt}]text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(text, return_tensors="pt").to(model.device) with torch.no_grad(): outputs = model.generate(**inputs, max_new_tokens=250, temperature=0.7)print(tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))~/.continue/config.json)json
{ "models": [ { "title": "OzyTalk 1.58b (Conversational RAM)", "provider": "openai", "model": "ozytalk-1.58b", "apiBase": "http://localhost:8766/v1", "apiKey": "sk-ozytalk-local-token" } ]}Desarrollado y mantenido por Angel Lopez (Xangel0s) como parte de la suite neuro-simbólica Ozzyvler / OzyBase.
import torchfrom peft import PeftModelfrom transformers import AutoModelForCausalLM, AutoTokenizer model_id = "Xangel0s/OzyTalk-1.58b"base_id = "unsloth/Qwen2.5-Coder-1.5B-Instruct-bnb-4bit" tokenizer = AutoTokenizer.from_pretrained(model_id)base_model = AutoModelForCausalLM.from_pretrained( base_id, torch_dtype=torch.bfloat16, device_map="auto")model = PeftModel.from_pretrained(base_model, model_id)model.eval() prompt = "Explica la diferencia entre procesos e hilos y cuando usar cada uno."messages = [ {"role": "system", "content": "Eres OzyTalk, un asistente de programacion amigable y preciso."}, {"role": "user", "content": prompt}]text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(text, return_tensors="pt").to(model.device) with torch.no_grad(): outputs = model.generate(**inputs, max_new_tokens=250, temperature=0.7)print(tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)){ "models": [ { "title": "OzyTalk 1.58b (Conversational RAM)", "provider": "openai", "model": "ozytalk-1.58b", "apiBase": "http://localhost:8766/v1", "apiKey": "sk-ozytalk-local-token" } ]}