GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
tayaee
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
tayaee
Model Tree
unsloth/Llama-3.1-8B-Instruct
Input Modalities
Output Modalities
Supported Functionality
GamblerInsight Smoke (gamblerinsight-1-llama-3.1-8b-smoke) is a smoke-test checkpoint for multi-hop
financial reasoning and a tool-calling agent. It is fine-tuned from
unsloth/Llama-3.1-8B-Instruct on the DGX Spark GB10 platform to validate the end-to-end
pipeline: 2-turn tool-calling trajectories, adapter merge into a standard
safetensors checkpoint, and HuggingFace Hub deployment for vLLM serving.
This release is intended for pipeline verification, not production trading.
The model can request two financial tools during reasoning:
lookup_stock_price: latest OHLCV snapshot for a ticker
(fields: ticker, close, open, high, low, volume, timestamp).lookup_technical_indicators: precomputed technical indicators for a ticker, including
RSI, MACD, BBANDS, SMA, EMA, ATR, OBV, MFI, divergence,
and priced_in_score.Example tool call:
json
{ "id": "call_01", "type": "function", "function": {"name": "lookup_stock_price", "arguments": "{\"ticker\": \"AAPL\"}"}}unsloth/Llama-3.1-8B-Instructgamblerinsight-1-llama-3.1-8b-smoketayaee/gamblerinsight-1-llama-3.1-8b-smokelora_bf16)r=8, alpha=16, dropout=0.05max_steps=10, learning_rate=0.0002, batch_size=2minimal_sft_smoke.jsonl (16 trajectories)0.7221 (average: 1.1611)lookup_stock_price,
lookup_technical_indicators), followed by a tool observation message.<think>...</think>,
a machine-readable forecast block in <forecast_json>...</forecast_json>,
and a human-readable final report section.bash
vllm serve tayaee/gamblerinsight-1-llama-3.1-8b-smoke \ --host 0.0.0.0 --port 8001 \ --served-model-name gamblerinsight-1-llama-3.1-8b-smoke \ --enable-auto-tool-choice \ --tool-call-parser hermesHealth check:
bash
curl -sf http://localhost:8001/healthcurl -sf http://localhost:8001/v1/modelspython
from openai import OpenAI client = OpenAI(base_url="http://localhost:8001/v1", api_key="EMPTY")tools = [ { "type": "function", "function": { "name": "lookup_stock_price", "description": "Look up the latest stock price snapshot for a ticker.", "parameters": { "type": "object", "properties": {"ticker": {"type": "string"}}, "required": ["ticker"], }, }, },]resp = client.chat.completions.create( model="gamblerinsight-1-llama-3.1-8b-smoke", messages=[{"role": "user", "content": "Analyze AAPL with RSI and MACD."}], tools=tools, tool_choice="auto",)print(resp.choices[0].message.content)python
from transformers import AutoModelForCausalLM, AutoTokenizer model_id = "tayaee/gamblerinsight-1-llama-3.1-8b-smoke"tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)model = AutoModelForCausalLM.from_pretrained( model_id, device_map="auto", trust_remote_code=True)messages = [{"role": "user", "content": "Analyze AAPL with RSI and MACD."}]prompt = tokenizer.apply_chat_template( messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(prompt, return_tensors="pt").to(model.device)outputs = model.generate(**inputs, max_new_tokens=512)print(tokenizer.decode(outputs[0], skip_special_tokens=True))unsloth/Llama-3.1-8B-Instruct.{ "id": "call_01", "type": "function", "function": {"name": "lookup_stock_price", "arguments": "{\"ticker\": \"AAPL\"}"}}vllm serve tayaee/gamblerinsight-1-llama-3.1-8b-smoke \ --host 0.0.0.0 --port 8001 \ --served-model-name gamblerinsight-1-llama-3.1-8b-smoke \ --enable-auto-tool-choice \ --tool-call-parser hermescurl -sf http://localhost:8001/healthcurl -sf http://localhost:8001/v1/modelsfrom openai import OpenAI client = OpenAI(base_url="http://localhost:8001/v1", api_key="EMPTY")tools = [ { "type": "function", "function": { "name": "lookup_stock_price", "description": "Look up the latest stock price snapshot for a ticker.", "parameters": { "type": "object", "properties": {"ticker": {"type": "string"}}, "required": ["ticker"], }, }, },]resp = client.chat.completions.create( model="gamblerinsight-1-llama-3.1-8b-smoke", messages=[{"role": "user", "content": "Analyze AAPL with RSI and MACD."}], tools=tools, tool_choice="auto",)print(resp.choices[0].message.content)from transformers import AutoModelForCausalLM, AutoTokenizer model_id = "tayaee/gamblerinsight-1-llama-3.1-8b-smoke"tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)model = AutoModelForCausalLM.from_pretrained( model_id, device_map="auto", trust_remote_code=True)messages = [{"role": "user", "content": "Analyze AAPL with RSI and MACD."}]prompt = tokenizer.apply_chat_template( messages, tokenize=False, add_generation_prompt=True)inputs = tokenizer(prompt, return_tensors="pt").to(model.device)outputs = model.generate(**inputs, max_new_tokens=512)print(tokenizer.decode(outputs[0], skip_special_tokens=True))