Alpaca
A fintech company providing stock brokerage services in an API-first manner, serving algorithmic traders and developers. The platform supports paper trading simulation accounts alongside live trading.
Timeline (last 90 days)
Alpaca and ShareGPT are popular instruction fine-tuning data formats, each with its own variants.
The trading bot only runs simulated trades, and the plan is to observe for a month before evaluating performance.
Common data format specifications include the Alpaca format and the ShareGPT format, both of which are supported for automatic parsing by mainstream training frameworks such as LLaMA-Factory and Axolotl.
Alpaca's official website is alpaca.markets, and newly opened paper trading accounts come with $100,000 in virtual funds.
The system pulls stock market data from Massive, calculates 12-month momentum scores across a universe of 50 stocks, and routes orders through SnapTrade to Alpaca's paper trading account for execution.
Notional ordering allows orders to be placed by dollar amount rather than share count; Alpaca supports notional orders and allows holding fractional shares, which is one of the reasons it was chosen as the execution layer.
When placing a notional order, the exact number of shares to be purchased is calculated by Alpaca at the real-time price at the time of execution, and the application itself is not responsible for this step.
After LLaMA was open-sourced, the community rapidly produced a large number of derivative models such as Alpaca and Vicuna.
Stanford University's Alpaca project in 2023 fine-tuned the LLaMA-7B model using 52,000 instruction data entries generated by GPT-3.5, at a cost of approximately $600.
Meta的Llama系列作为开源模型对齐程度较轻，催生了Vicuna、Alpaca等数以千计的衍生模型
1 more timeline events
All Facts (10)
Stanford's Alpaca project used GPT-3.5 to generate 52K instruction data points for training a 7B-parameter LLaMA.
60%UnverifiedAlpaca and ShareGPT are popular instruction fine-tuning data formats, each with its own variants.
50%UnverifiedCommon data format specifications include the Alpaca format and the ShareGPT format, both of which are supported for automatic parsing by mainstream training frameworks such as LLaMA-Factory and Axolotl.
50%UnverifiedThe system pulls stock market data from Massive, calculates 12-month momentum scores across a universe of 50 stocks, and routes orders through SnapTrade to Alpaca's paper trading account for execution.
50%UnverifiedAlpaca's official website is alpaca.markets, and newly opened paper trading accounts come with $100,000 in virtual funds.
50%UnverifiedNotional ordering allows orders to be placed by dollar amount rather than share count; Alpaca supports notional orders and allows holding fractional shares, which is one of the reasons it was chosen as the execution layer.
50%UnverifiedWhen placing a notional order, the exact number of shares to be purchased is calculated by Alpaca at the real-time price at the time of execution, and the application itself is not responsible for this step.
50%UnverifiedAfter LLaMA was open-sourced, the community rapidly produced a large number of derivative models such as Alpaca and Vicuna.
50%UnverifiedStanford University's Alpaca project in 2023 fine-tuned the LLaMA-7B model using 52,000 instruction data entries generated by GPT-3.5, at a cost of approximately $600.
50%UnverifiedMeta的Llama系列作为开源模型对齐程度较轻，催生了Vicuna、Alpaca等数以千计的衍生模型
50%