-
### 🚀 The feature, motivation and pitch
Hi,
I have
- provisioned vm with A100
- installed needed driver.
- downloaded Llama 3.1 8B Instruct model
- llama build, configure & run (using meta-ref…
-
Currently, many evaluations of long text models reference LongBench results. However, n-gram based metrics do not truly reflect the quality of responses. Many papers have adopted the method of using G…
-
### ENDGAME ISSUE - PHASE III
### Overview
Implement practical AI-powered features to enhance user experience, improve content quality, and increase engagement while maintaining reasonable computa…
-
### 🚀 The feature, motivation and pitch
I insist on trying to learn and study the llama-stack API and its examples, but they are complex. There’s no documentation, or I’m unable to access it.
I wa…
-
### Objective
Develop a robust AI Judging System to evaluate code submissions quickly and objectively in the AI Hacking League.
### Tasks
1. **API Integration:**
- Integrate OpenAI's GPT-4 model t…
-
Hi, I have a question about the tokenizer mismatch.
When the reference model is fixed to be "gpt-j-6B", several scoring models do not share the same tokenizer, such as "gpt-neox-20b" and "llama". …
-
Items to draft/develop:
- changelog DB table/model. Possibility to use triggers, or some logic in a python script with django models.
- changelog file to be generated from DB at release and placed in…
fyvon updated
3 weeks ago
-
### Willingness to contribute
No. I cannot contribute this feature at this time.
### Proposal Summary
The 2.12.2 release added support for the `predict_stream` method for custom Python models. Whil…
-
Need to implement a method to calculate PGS that only have the `dosage_#_weight`-type columns in the [schema](https://www.pgscatalog.org/downloads/#scoring_columns).
### Discussed in https://github…
-
Integrating HuggingFace models scoring function iVSM in vantage. We have the jar files, and the steps needed. We need help in integrating it in the image.